Power patrol infrared feature decoupling and hidden danger absolute temperature rise automatic identification method, device, equipment and medium

Abstract

This application provides a method, device, electronic equipment, and medium for decoupling infrared features and automatically identifying absolute temperature rise of potential hazards during power equipment inspection. The method involves acquiring the original multi-view infrared thermal image of the target power equipment and a pre-constructed three-dimensional structure, including the structural spatial coordinates and material thermal conductivity properties. The original multi-view infrared thermal image is then pose-normalized to obtain the multi-view infrared thermal image data source. A pre-configured environmental radiation elimination matrix is ​​acquired and used to decouple the features of the multi-view infrared thermal image data source, generating an infrared feature map of the equipment and a surface temperature mapping feature set. The feature set is fused with the three-dimensional structure to determine thermal boundary conditions and perform heat conduction simulation, constructing a three-dimensional temperature field distribution map. After locating candidate heating areas and performing fault confirmation processing, absolute heat quantification is performed to determine the target heat dissipation and absolute temperature rise parameters, and the identification results are output. This application demonstrates high reliability in hazard identification.

Description

Technical Field

[0001] This application relates to the field of intelligent inspection and testing technology for power equipment. Specifically, it relates to a method, device, electronic equipment, and computer-readable storage medium for decoupling infrared features and automatically identifying the absolute temperature rise of potential hazards during power equipment inspection. Background Technology

[0002] With the continuous development of new power systems, the types and quantities of power equipment are increasing dramatically, making real-time monitoring of their operating status crucial for power grid safety. Infrared thermal imaging, as a non-contact detection method, is widely used in power line inspections to detect abnormal heating in equipment. In complex outdoor inspection scenarios, infrared thermal images of power equipment are often affected by environmental radiation, varying viewing angles, and the complex spatial structure of the equipment itself. This requires relevant detection solutions to effectively isolate external interference and accurately reconstruct the true internal heating conditions of the equipment by combining its physical form.

[0003] Existing infrared thermal imaging hazard detection schemes for power equipment typically employ a detection mechanism based on pixel-level threshold segmentation of two-dimensional images. This scheme first acquires two-dimensional infrared images of the equipment using an infrared camera at a fixed angle, and then transforms the image's grayscale matrix into a two-dimensional planar temperature matrix using a globally unified empirical temperature calibration formula. Subsequently, image morphology processing methods are used to directly extract the local highest temperature pixels from this two-dimensional temperature matrix as the heat generation center. Finally, the extracted two-dimensional highest temperature value is rigidly compared with the equipment's preset safe operating temperature limit, thereby generating and outputting an overheating hazard alarm for the equipment.

[0004] However, this detection scheme based on pixel-level segmentation of two-dimensional images has significant technical drawbacks. Because it performs global temperature calibration and extraction directly at the two-dimensional image level, it cannot isolate the complex coupling interference between outdoor environmental radiation and the equipment's own heat generation, resulting in extracted temperature data mixed with a large amount of environmental thermal noise. Furthermore, relying solely on two-dimensional plane pixels for temperature comparison completely ignores the actual three-dimensional spatial structure and material thermal conductivity characteristics of the power equipment. It cannot trace the heat conduction path and attenuation diffusion process of the internal heat source, thus failing to provide a deep quantitative assessment of actual heat generation and absolute temperature rise. When facing complex equipment or deep internal hazards, it is prone to high false alarm and false negative rates, failing to meet the demands of modern power distribution networks for accurate hazard location. Summary of the Invention

[0005] This application provides a method, device, electronic device, and computer-readable storage medium for decoupling infrared features and automatically identifying the absolute temperature rise of potential hazards during power inspections, in order to at least alleviate the aforementioned technical problems.

[0006] A method for automatic identification of hidden dangers by decoupling infrared features and absolute temperature rise during power line inspection includes the following steps: The original multi-view infrared thermal image of the target power equipment and the pre-constructed three-dimensional structure of the target power equipment are obtained. The three-dimensional structure of the target power equipment includes the structural spatial coordinates and material thermal conductivity properties of the target power equipment. The original multi-view infrared thermal image is subjected to pose normalization processing to obtain the multi-view infrared thermal image data source. A pre-configured environmental radiation rejection matrix is ​​obtained, and the multi-view infrared thermal image data source is decoupled using the environmental radiation rejection matrix to generate an infrared feature map of the target power equipment. A surface temperature mapping feature set of the target power equipment is then generated based on the infrared feature map of the target power equipment. The surface temperature mapping feature set of the target power equipment is structurally fused with the three-dimensional structure of the target power equipment to determine the thermal boundary conditions of the target power equipment. Based on the thermal boundary conditions of the target power equipment, heat conduction is deduced to construct a three-dimensional temperature field distribution map of the target power equipment. Candidate heating areas are located from the three-dimensional temperature field distribution map of the target power equipment. Fault confirmation processing is performed on the candidate heating areas to generate candidate heating area fault confirmation results. After the candidate heating area passes the fault confirmation processing, the absolute heat of the candidate heating area is quantified to determine the target heat consumption and the absolute temperature rise parameter of the candidate heating area. Based on the target heat consumption and the absolute temperature rise parameter of the candidate heating area, the hazard absolute temperature rise identification result is output.

[0007] Optionally, the step of performing pose normalization processing on the original multi-view infrared thermal image to obtain a multi-view infrared thermal image data source includes: Obtain the shooting pose parameters corresponding to the original multi-view infrared thermal image; Based on the shooting pose parameters, the original multi-view infrared thermal image is spatially projected and transformed, and the original multi-view infrared thermal image after spatial projection transformation is contour aligned according to the physical geometric contour in the three-dimensional structure of the target power equipment to obtain a pose-normalized thermal image sequence. The thermal intensity of the pose-normalized thermal image sequence is calibrated to obtain the multi-view infrared thermal image data source.

[0008] Optionally, the step of performing thermal intensity calibration on the pose-normalized thermal image sequence to obtain the multi-view infrared thermal image data source includes: Obtain the grayscale mapping features of pixels in the pose-normalized heatmap sequence; The grayscale mapping features are subjected to temperature inversion transformation according to the preset blackbody radiation reference calibration mapping relationship to obtain an absolute temperature distribution matrix sequence. The absolute temperature distribution matrix sequence is linked with the pose-normalized thermal image sequence to obtain the multi-view infrared thermal image data source.

[0009] Optionally, the step of using the environmental radiation removal matrix to decouple features from the multi-view infrared thermal image data source and generate an infrared feature map of the target power equipment includes: Obtain the pre-configured environmental radiation rejection matrix, wherein the pre-configured environmental radiation rejection matrix is ​​determined by the correspondence between the background radiation characteristics of the inspection environment and the infrared imaging channel, and the background radiation characteristics of the inspection environment and the correspondence between the infrared imaging channel jointly define the distribution of matrix elements in the pre-configured environmental radiation rejection matrix; The environmental radiation removal matrix is ​​applied to the multi-view infrared thermal image data source to establish an independent component analysis matrix. The energy weight distribution ratio of different heat source components is determined by performing blind source separation iterative calculation on the multi-view infrared thermal image data source using the independent component analysis matrix. The environmental interference component is determined based on the energy weight distribution ratio, and the environmental interference component is removed from the multi-view infrared thermal image data source to generate the infrared feature map of the target power equipment.

[0010] Optionally, the step of generating a surface temperature mapping feature set of the target power equipment based on the infrared feature map of the target power equipment includes: Determine the thermal gradient features in the infrared feature map of the target power equipment; Obtain a preset temperature threshold that matches the device type and operating conditions of the target power equipment, wherein the preset temperature threshold is pre-configured based on the device type and operating conditions of the target power equipment; Spatial connectivity analysis is performed on the thermal gradient features to delineate target temperature connectivity regions with temperature values ​​greater than the preset temperature threshold. The contour boundary coordinates corresponding to the target temperature connected region are determined, and the contour boundary coordinates are structurally recombined with the infrared feature map of the target power equipment to generate the surface temperature mapping feature set of the target power equipment.

[0011] Optionally, the step of structurally fusing the surface temperature mapping feature set of the target power equipment with the three-dimensional structure of the target power equipment to determine the thermal boundary conditions of the target power equipment includes: Obtain a preset physical depth level pre-configured according to the entity structure layer relationship in the three-dimensional structure of the target power equipment, wherein the entity structure layer relationship is used to define the layer position of the preset physical depth level in the three-dimensional structure of the target power equipment; The surface temperature mapping feature set of the target power equipment is hierarchically parsed according to the preset physical depth level in order to extract the surface heat dissipation gradient grid representing different preset physical depth levels. The surface heat dissipation gradient grid is mapped to the three-dimensional structure of the target power equipment in spatial coordinates to determine the three-dimensional spatial temperature distribution profile. The thermal boundary conditions of the target power equipment are determined based on the three-dimensional spatial temperature distribution profile.

[0012] Optionally, the step of constructing a three-dimensional temperature field distribution map of the target power equipment by performing heat conduction deduction based on the thermal boundary conditions of the target power equipment includes: The set of unsteady-state heat conduction inference rules is called to match the material thermal conductivity properties in the three-dimensional structure of the target power equipment and the thermal boundary conditions of the target power equipment. The set of unsteady-state heat conduction inference rules includes rules for heat transfer along the material thermal conductivity direction, rules for heat diffusion at the junction of structural layers, and rules for determining the attenuation of heat with spatial distance. Using the unsteady-state heat conduction deduction rule set, reverse heat conduction deduction is performed based on the thermal boundary conditions of the target power equipment to reconstruct the internal heat conduction path; Temperature field reconstruction is performed along the internal heat conduction path to extract the geometric features of the three-dimensional isothermal surface; Based on the geometric features of the three-dimensional isothermal surface, a three-dimensional spatial structure is encapsulated to construct a three-dimensional temperature field distribution map of the target power equipment.

[0013] Optionally, the step of performing fault confirmation processing on the candidate heating area and generating fault confirmation results for the candidate heating area includes: The preset range is determined based on the spatial adjacency relationship between the candidate heat-generating area and adjacent structures in the three-dimensional structure of the target power equipment; Track the heat decay curve of the candidate heat-generating region radiating to adjacent structures within the preset range; A trend fitting analysis is performed on the heat decay curve to determine the heat generation trend characteristics; Obtain a pre-constructed feature set of non-faulty environment heat sources based on non-faulty environment heat source samples; The heat generation trend features are compared with the heat source feature set of the non-faulty environment to prevent false alarms, and the fault confirmation results of the candidate heat generation area are generated based on the false alarm prevention cross-comparison results.

[0014] Optionally, the step of quantifying the absolute heat of the candidate heating region to determine the target heat dissipation and the absolute temperature rise parameter of the candidate heating region includes: Determine the current-carrying cross-sectional area parameters and estimated operating current values ​​of the candidate heating region in the three-dimensional structure of the target power equipment; Based on the current-carrying cross-sectional area parameter and the estimated operating current value, an internal resistance heating quantification rule is established. The internal resistance heating quantification rule includes a judgment rule that the smaller the current-carrying cross-sectional area, the higher the resistance heating contribution; a judgment rule that the larger the estimated operating current value, the higher the resistance heating contribution; and a quantification rule corresponding to the heat accumulation state and heat dissipation. Discrete sampling is performed on the temperature field corresponding to the candidate heating region in the three-dimensional temperature field distribution map of the target power equipment to obtain discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region. The discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region are substituted into the internal resistance heating quantification rule for thermoelectric coupling iterative matching processing to determine the target heating power consumption. The absolute temperature rise parameter of the candidate heating region is determined based on the discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region and the thermal boundary conditions of the target power equipment.

[0015] A power line inspection infrared feature decoupling and hazard absolute temperature rise automatic identification device includes a pose normalization processing module, an infrared feature decoupling module, a three-dimensional temperature field construction module, and a hazard absolute temperature rise identification result output module. The pose normalization processing module is configured to acquire the original multi-view infrared thermal image of the target power equipment and the pre-constructed three-dimensional structure of the target power equipment. The three-dimensional structure of the target power equipment includes the structural space coordinates and material thermal conductivity properties of the target power equipment. The pose normalization processing is performed on the original multi-view infrared thermal image to obtain the multi-view infrared thermal image data source. The infrared feature decoupling module is configured to obtain a pre-configured environmental radiation elimination matrix, use the environmental radiation elimination matrix to decouple the features of the multi-view infrared thermal image data source, generate an infrared feature map of the target power equipment, and generate a surface temperature mapping feature set of the target power equipment based on the infrared feature map of the target power equipment. The three-dimensional temperature field construction module is configured to structurally fuse the surface temperature mapping feature set of the target power equipment with the three-dimensional structure of the target power equipment, determine the thermal boundary conditions of the target power equipment, perform heat conduction deduction based on the thermal boundary conditions of the target power equipment, and construct a three-dimensional temperature field distribution map of the target power equipment. The hidden danger absolute temperature rise identification result output module is configured to locate candidate heating areas from the three-dimensional temperature field distribution map of the target power equipment, perform fault confirmation processing on the candidate heating areas, generate a fault confirmation result for the candidate heating areas, and after the candidate heating areas pass the fault confirmation processing, perform absolute heat quantification on the candidate heating areas to determine the target heat consumption and the absolute temperature rise parameter of the candidate heating areas, and output the hidden danger absolute temperature rise identification result based on the target heat consumption and the absolute temperature rise parameter of the candidate heating areas.

[0016] An electronic device includes: a memory, a processor, and an automatic identification program for decoupling infrared features and identifying absolute temperature rise of potential hazards during power patrols, stored in the memory and executable on the processor. The automatic identification program for decoupling infrared features and identifying absolute temperature rise of potential hazards during power patrols is configured to implement the steps of the above-described method for automatic identification of decoupling infrared features and identifying absolute temperature rise of potential hazards during power patrols.

[0017] A computer-readable storage medium stores a program for automatic identification of absolute temperature rise and decoupling of infrared features during power inspections, wherein the program, when executed by a processor, implements the steps of the aforementioned method for automatic identification of absolute temperature rise and decoupling of infrared features during power inspections.

[0018] The technical advantages of the technical solution provided in this application are: This application presents a method for decoupling infrared features and automatically identifying absolute temperature rise of potential hazards during power line inspections. Addressing the shortcomings of traditional two-dimensional image pixel-level segmentation schemes, which fail to remove environmental radiation coupling interference and result in extracted temperature data being contaminated with significant environmental thermal noise, this method obtains a pre-configured environmental radiation removal matrix. This matrix is ​​then used to decouple features from multi-view infrared thermal image data sources, generating an infrared feature map of the target power equipment and further generating a surface temperature mapping feature set. This solves the problem of severe external noise interference in infrared data in traditional schemes. Compared to the traditional approach of performing global empirical calibration at the two-dimensional image level, this application introduces independent matrix operations to remove radiation components. This effectively removes environmental interference in complex outdoor scenarios, allowing the extracted surface temperature mapping feature set to more purely represent the infrared information of the equipment itself. This ensures the anti-interference capability and reliability of subsequent temperature rise determination features at the underlying data source.

[0019] Furthermore, addressing the shortcomings of traditional solutions that detach from the actual three-dimensional spatial structure and material thermal conductivity characteristics of power equipment, and are unable to trace heat conduction paths for quantifying heat generation and power consumption, this application structurally fuses the surface temperature mapping feature set of the target power equipment with the three-dimensional structure of the target power equipment, including structural spatial coordinates and material thermal conductivity properties. This determines the thermal boundary conditions of the target power equipment, and based on this, a three-dimensional temperature field distribution map of the target power equipment is constructed through heat conduction deduction. This solves the problem of the traditional solution relying solely on planar features, resulting in a lack of understanding of the internal thermodynamic state. Compared to the isolated planar maximum temperature extraction of traditional solutions, this application combines the three-dimensional coordinates of the entity with material properties for a reverse thermodynamic deduction from the surface inwards. This expands the perception of equipment temperature from a thin two-dimensional plane to a complete three-dimensional space, better reconstructing the internal heat diffusion topology of the equipment.

[0020] Based on this, this application locates candidate heat-generating areas from the constructed three-dimensional temperature field distribution map. After fault confirmation processing, the absolute heat of these areas is quantified to determine the target heat dissipation and the absolute temperature rise parameters of the candidate heat-generating areas, thereby outputting the final absolute temperature rise identification result of the potential hazard. Compared with the traditional solution that only makes a hard comparison between the surface temperature value and the safety limit, this application relies on the three-dimensional spatial structure to deeply explore the actual heat dissipation of the underlying layer for absolute heat assessment. This eliminates misjudgments caused by surface reflection or non-structural heat accumulation, resulting in a lower false alarm rate and false negative rate when dealing with complex electrical equipment and highly concealed internal faults. This makes the output hazard identification result more valuable for judging overheating hazards under complex operating conditions. Attached Figure Description

[0021] Figure 1 This application provides an embodiment of a scenario for decoupling infrared features during power line inspections and automatically identifying the absolute temperature rise of potential hazards. Figure 2 This application provides an embodiment of a method for decoupling infrared features during power line inspections and automatically identifying the absolute temperature rise of potential hazards. Figure 3 This application provides an embodiment of a power inspection infrared feature decoupling and hazard absolute temperature rise automatic identification device; Figure 4 An electronic device is described in an embodiment of this application; Figure 5 This is a computer-readable storage medium according to an embodiment of the present application. Detailed Implementation

[0022] like Figure 1 As shown, this is an embodiment of the present application of a scenario for automatic identification of absolute temperature rise and decoupling of infrared features during power line inspection. Figure 2As shown in the figure, this application provides an embodiment of a method for decoupling infrared features during power line inspections and automatically identifying the absolute temperature rise of potential hazards. The method includes the following steps: The original multi-view infrared thermal image of the target power equipment and the pre-constructed three-dimensional structure of the target power equipment are obtained. The three-dimensional structure of the target power equipment includes the structural spatial coordinates and material thermal conductivity properties of the target power equipment. The original multi-view infrared thermal image is subjected to pose normalization processing to obtain the multi-view infrared thermal image data source. A pre-configured environmental radiation rejection matrix is ​​obtained, and the multi-view infrared thermal image data source is decoupled using the environmental radiation rejection matrix to generate an infrared feature map of the target power equipment. A surface temperature mapping feature set of the target power equipment is then generated based on the infrared feature map of the target power equipment. The surface temperature mapping feature set of the target power equipment is structurally fused with the three-dimensional structure of the target power equipment to determine the thermal boundary conditions of the target power equipment. Based on the thermal boundary conditions of the target power equipment, heat conduction is deduced to construct a three-dimensional temperature field distribution map of the target power equipment. Candidate heating areas are located from the three-dimensional temperature field distribution map of the target power equipment. Fault confirmation processing is performed on the candidate heating areas to generate candidate heating area fault confirmation results. After the candidate heating area passes the fault confirmation processing, the absolute heat of the candidate heating area is quantified to determine the target heat consumption and the absolute temperature rise parameter of the candidate heating area. Based on the target heat consumption and the absolute temperature rise parameter of the candidate heating area, the hazard absolute temperature rise identification result is output.

[0023] Optionally, the step of performing pose normalization processing on the original multi-view infrared thermal image to obtain a multi-view infrared thermal image data source includes: Obtain the shooting pose parameters corresponding to the original multi-view infrared thermal image; Based on the shooting pose parameters, the original multi-view infrared thermal image is spatially projected and transformed, and the original multi-view infrared thermal image after spatial projection transformation is contour aligned according to the physical geometric contour in the three-dimensional structure of the target power equipment to obtain a pose-normalized thermal image sequence. The thermal intensity of the pose-normalized thermal image sequence is calibrated to obtain the multi-view infrared thermal image data source.

[0024] Preferably, when performing pose normalization processing on the original multi-view infrared thermal image, the shooting pose parameters are first read from the infrared acquisition record corresponding to the original multi-view infrared thermal image. The infrared acquisition record is the imaging association data synchronously saved during the acquisition process of the original multi-view infrared thermal image. The imaging association data in the infrared acquisition record is used to define the spatial observation relationship between the original multi-view infrared thermal image and the target power equipment. The shooting pose parameters include the imaging position, imaging orientation, lens field of view, shooting distance, and infrared thermal image acquisition time of the infrared acquisition device relative to the target power equipment. Among them, the imaging position expresses the spatial orientation of the infrared acquisition device relative to the target power equipment when acquiring the original multi-view infrared thermal image; the imaging orientation expresses the correspondence between the optical axis orientation of the infrared acquisition device and the visible outer surface of the target power equipment; the lens field of view expresses the coverage area of ​​the visible outer surface of the target power equipment in the original multi-view infrared thermal image; the shooting distance expresses the spatial interval between the infrared acquisition device and the visible outer surface of the target power equipment; and the infrared thermal image acquisition time expresses the acquisition sequence relationship between the original multi-view infrared thermal images under different perspectives. By using the shooting pose parameters, each original multi-view infrared thermal image can be transformed from a simple two-dimensional thermal image into an original multi-view infrared thermal image with an imaging source and spatial observation relationship. This allows subsequent spatial projection transformation to determine the projection relationship between each original multi-view infrared thermal image and the three-dimensional structure of the target power equipment based on the shooting pose parameters.

[0025] Preferably, before participating in pose normalization processing, the three-dimensional structure of the target power equipment is first represented by a readable spatial structure based on its structural spatial coordinates and material thermal conductivity properties. The structural spatial coordinates record the coordinate distribution of different structural positions on the outer surface of the target power equipment under the same spatial reference, and the material thermal conductivity properties record the thermal conductivity differences between the outer surface of the target power equipment and its adjacent structural layers during heat transfer. Further, a physical geometric contour is extracted from the three-dimensional structure of the target power equipment based on the structural spatial coordinates. This physical geometric contour includes the outer surface boundary of the target power equipment, structural transition boundaries, and a visible installation contour. The outer surface boundary of the target power equipment defines the outer perimeter of the visible outer surface, the structural transition boundaries define the transition positions between different structural surfaces on the visible outer surface, and the visible installation contour defines the contour position corresponding to the installation structure on the visible outer surface. The physical geometric contour is not extracted in isolation from the original multi-view infrared thermal image, but rather originates from the structural space coordinates in the three-dimensional structure of the target power equipment. The physical geometric contour continues to serve as a geometric reference for contour alignment when performing spatial projection transformation thermal image, so that subsequent contour alignment no longer depends solely on the infrared intensity boundary in the two-dimensional thermal image, but establishes a correspondence with the actual external shape structure expressed by the three-dimensional structure of the target power equipment.

[0026] Preferably, the spatial projection transformation specifically involves: determining the imaging ray corresponding to each pixel in each original multi-view infrared thermal image based on the imaging position, imaging orientation, lens field of view, and shooting distance in the shooting pose parameters; the imaging ray is used to express the spatial observation direction of the pixel in the original multi-view infrared thermal image from the infrared acquisition device to the visible outer surface of the target power equipment; then, based on the intersection relationship between the imaging ray and the structural spatial coordinates in the three-dimensional structure of the target power equipment, determining the corresponding position of the pixel in the original multi-view infrared thermal image on the outer surface of the target power equipment in the three-dimensional structure of the target power equipment. The corresponding position on the outer surface of the target power equipment is further matched with the physical geometric contour, so that the boundary of the outer surface of the target power equipment, the thermal distribution area of ​​the outer surface of the target power equipment, and the visible outer surface of the target power equipment in each original multi-view infrared thermal image can be projected onto the same three-dimensional structure of the target power equipment. After the spatial projection transformation, the original multi-view infrared thermal images, which originally had differences in perspective, distance, and appearance scale under different shooting angles, are converted into spatial projection transformation thermal images with a common structural reference. The spatial projection transformation thermal images continue to participate in contour alignment to eliminate the positional offset caused by different shooting angles on the corresponding positions on the outer surface of the same target power equipment.

[0027] Preferably, when aligning the spatial projection transformation heatmap according to the physical geometric contour of the target power equipment's three-dimensional structure, the outer surface boundary of the target power equipment in the spatial projection transformation heatmap is first matched with the outer surface boundary of the target power equipment in the physical geometric contour. Then, the structural transition region in the spatial projection transformation heatmap is matched with the structural transition boundary in the physical geometric contour, and the visible installation region in the spatial projection transformation heatmap is matched with the visible installation contour in the physical geometric contour. Subsequently, based on the matching relationships of the outer surface boundary of the target power equipment, the structural transition boundary, and the visible installation contour, the spatial projection transformation heatmap undergoes translation correction, orientation correction, and scale resampling. After translation correction, orientation correction, and scale resampling, the spatial projection transformation heatmap can express the temperature distribution at corresponding positions on the outer surface of the same target power equipment under the spatial reference of the same three-dimensional structure, thus forming a contour-aligned heatmap. The contour-aligned thermal images are further arranged according to the infrared thermal image acquisition time and the corresponding position on the outer surface of the target power equipment, thereby obtaining a pose-normalized thermal image sequence. This allows different contour-aligned thermal images in the pose-normalized thermal image sequence to be subsequently calibrated for thermal intensity using a unified physical geometric contour as a reference.

[0028] Preferably, during the formation of the pose-normalized thermal image sequence, the overlapping observation relationships of corresponding positions on the outer surface of the same target power equipment under different viewpoints are also organized. These overlapping observation relationships originate from the corresponding positions on the outer surface of the target power equipment corresponding to pixels in the spatial projection transformation thermal images. When multiple spatial projection transformation thermal images are mapped to the corresponding positions on the outer surface of the same target power equipment within the three-dimensional structure of the target power equipment, the visibility of these positions under different viewpoints is first determined based on the imaging orientation, lens field of view, and shooting distance in the shooting pose parameters. Then, based on the visibility, contour-aligned thermal images with lower projection distortion are selected from the multiple spatial projection transformation thermal images to participate in the organization of the pose-normalized thermal image sequence. Through this processing method, the pose-normalized thermal image sequence is not merely a simple sorting of the original multi-view infrared thermal images, but rather a rebinding of the thermal information in the original multi-view infrared thermal images to the corresponding positions on the outer surface of the target power equipment within the three-dimensional structure of the target power equipment. This allows subsequent thermal intensity calibration to express temperature around the corresponding positions on the outer surface of the same target power equipment, reducing surface temperature mismatches caused by viewpoint switching.

[0029] Preferably, after the pose-normalized heatmap sequence is formed, the pose-normalized heatmap sequence is subjected to thermal intensity calibration. The thermal intensity calibration first reads the grayscale mapping features of each contour-aligned heatmap in the pose-normalized heatmap sequence. The grayscale mapping features express the correspondence between the grayscale level of a pixel in the contour-aligned heatmap and the infrared radiation response. Further, the grayscale mapping features are subjected to temperature inversion transformation according to a preset blackbody radiation reference calibration mapping relationship. The blackbody radiation reference calibration mapping relationship is pre-configured based on the response record of the infrared acquisition device to the blackbody radiation source under calibration conditions. The response record of the blackbody radiation source is used to express the grayscale level change relationship formed by the infrared acquisition device under different known thermal radiation conditions. The blackbody radiation reference calibration mapping relationship transforms the grayscale mapping features into an absolute temperature distribution matrix sequence that can express the meaning of physical temperature based on the response record of the blackbody radiation source. The absolute temperature distribution matrix sequence continues to be feature-associated with the pose-normalized thermal image sequence, so that the pixel position in each contour-aligned thermal image, the corresponding position on the outer surface of the target power equipment, and the temperature value in the absolute temperature distribution matrix sequence form a one-to-one technical relationship. This one-to-one technical relationship continues to participate in the generation of multi-view infrared thermal image data sources.

[0030] Preferably, the thermal intensity calibration differs from the traditional approach of directly applying a globally unified temperature conversion to the two-dimensional infrared image. In this application, the thermal intensity calibration is based on the pose-normalized thermal image sequence. First, the geometric perspective is unified using the shooting pose parameters and the three-dimensional structure of the target power equipment. Then, temperature inversion transformation is performed on the geometrically unified pose-normalized thermal image sequence. Since the different contour-aligned thermal images in the pose-normalized thermal image sequence have been aligned according to the physical geometric contours, the thermal intensity calibration can convert the thermal intensity changes of corresponding positions on the outer surface of the same target power equipment under different perspectives into a temperature expression under the same spatial reference. Therefore, the absolute temperature distribution matrix sequence is not an isolated two-dimensional temperature table, but a temperature expression associated with the corresponding position on the outer surface of the target power equipment in the three-dimensional structure of the target power equipment. This temperature expression continues to participate in the generation of the multi-view infrared thermal image data source, so as to perform feature decoupling on the multi-view infrared thermal image data source subsequently.

[0031] Preferably, in an application scenario, when the target power equipment is a power distribution device with an outer surface of a housing, an outer surface of a heat sink, and an outer surface of a wiring end, the physical geometric contour can be composed of the outer surface boundary of the housing, the outer surface boundary of the heat sink, and the outer surface boundary of the wiring end in the three-dimensional structure of the target power equipment. The outer surface boundary of the housing is used to express the outer perimeter of the outer surface of the housing within the three-dimensional structure of the target power equipment; the outer surface boundary of the heat sink is used to express the outer perimeter of the outer surface of the heat sink within the three-dimensional structure of the target power equipment; and the outer surface boundary of the wiring end is used to express the outer perimeter of the outer surface of the wiring end within the three-dimensional structure of the target power equipment. When the original multi-view infrared thermal images are acquired from the front of the outer surface of the housing, the side of the outer surface of the heat sink, and the oblique direction of the outer surface of the wiring end, the size and shape of each original multi-view infrared thermal image will differ in the two-dimensional image. At this point, the original multi-view infrared thermal images are first projected onto the outer surface of the casing, the outer surface of the heat sink, and the outer surface of the wiring terminals in the three-dimensional structure of the target power equipment using the shooting pose parameters, to form spatial projection transformation thermal images corresponding to the outer surface of the casing, the outer surface of the heat sink, and the outer surface of the wiring terminals, respectively. Then, the spatial projection transformation thermal images are aligned according to the physical geometric contours to form a pose-normalized thermal image sequence that can express the temperature distribution of the corresponding outer surfaces of the casing, the heat sink, and the wiring terminals. After thermal intensity calibration, the pose-normalized thermal image sequence forms a multi-view infrared thermal image data source, enabling subsequent processing to distinguish the true temperature distribution of the target power equipment's outer surface from the image differences caused by changes in the shooting angle based on the multi-view infrared thermal image data source.

[0032] Preferably, the multi-view infrared thermal image data source is jointly formed by the relationship between the pose-normalized thermal image sequence, the absolute temperature distribution matrix sequence, the shooting pose parameters, and the three-dimensional structure of the target power equipment. Specifically, the pose-normalized thermal image sequence represents infrared thermal images with contour alignment completed from different viewpoints; the absolute temperature distribution matrix sequence represents the absolute temperature corresponding to the position on the outer surface of each target power equipment in the pose-normalized thermal image sequence; the shooting pose parameters represent the imaging source of each original multi-view infrared thermal image; and the three-dimensional structure of the target power equipment represents the position on the outer surface of the target power equipment to which the aforementioned absolute temperature is attached. After the above association, the multi-view infrared thermal image data source can simultaneously retain the thermal intensity information of the infrared thermal image, the spatial structure information of the three-dimensional structure of the target power equipment, and the viewpoint source information of the shooting pose parameters. The multi-view infrared thermal image data source continues to serve as input when performing feature decoupling for the environmental radiation culling matrix, enabling the subsequently generated infrared feature map of the target power equipment to be built on the data that has completed viewpoint unification and thermal intensity calibration.

[0033] Optionally, the step of performing thermal intensity calibration on the pose-normalized thermal image sequence to obtain the multi-view infrared thermal image data source includes: Obtain the grayscale mapping features of pixels in the pose-normalized heatmap sequence; The grayscale mapping features are subjected to temperature inversion transformation according to the preset blackbody radiation reference calibration mapping relationship to obtain an absolute temperature distribution matrix sequence. The absolute temperature distribution matrix sequence is linked with the pose-normalized thermal image sequence to obtain the multi-view infrared thermal image data source.

[0034] Preferably, when calibrating the thermal intensity of the pose-normalized heatmap sequence, the pixel arrangement relationship of each contour-aligned heatmap in the pose-normalized heatmap sequence is first read, and the pixel arrangement relationship is then correlated with the corresponding position on the outer surface of the target power equipment formed in the aforementioned pose normalization process. The pixel arrangement relationship expresses the order of pixels in the row and column directions in the contour-aligned heatmap, and the corresponding position on the outer surface of the target power equipment expresses the spatial attachment position of the pixel within the three-dimensional structure of the target power equipment. By correlating the pixel arrangement relationship with the corresponding position on the outer surface of the target power equipment, the corresponding position on the outer surface of the target power equipment for each pixel in the pose-normalized heatmap sequence can be determined. This ensures that the grayscale mapping features subsequently obtained from the pose-normalized heatmap sequence are no longer two-dimensional image grayscale information detached from the three-dimensional structure of the target power equipment, but rather an infrared radiation response expression spatially correlated with the corresponding position on the outer surface of the target power equipment. This infrared radiation response expression continues to participate in subsequent temperature inversion conversion.

[0035] Preferably, the grayscale mapping feature includes the grayscale level of a pixel, the pixel position in the contour-aligned heatmap, the corresponding position on the outer surface of the target power equipment, and the change relationship of the grayscale level with the infrared heatmap acquisition time. The grayscale level expresses the intensity of the infrared radiation response received by the infrared acquisition device at the corresponding pixel position; the pixel position expresses the two-dimensional arrangement position of the grayscale level in the contour-aligned heatmap; the corresponding position on the outer surface of the target power equipment expresses the spatial attachment position of the grayscale level within the three-dimensional structure of the target power equipment; and the change relationship of the grayscale level with the infrared heatmap acquisition time expresses the difference in infrared radiation response at the corresponding position on the outer surface of the same target power equipment under different viewing angles and acquisition sequences. Thus, the grayscale mapping feature simultaneously retains the source of the grayscale level in the contour-aligned heatmap, the source of the corresponding position on the outer surface of the target power equipment within the three-dimensional structure of the target power equipment, and the source of the viewing angle corresponding to the infrared heatmap acquisition time. The grayscale mapping feature continues to serve as the processing basis for temperature inversion conversion, converting the intensity of the infrared radiation response in the grayscale mapping feature into the absolute temperature value of the pixel.

[0036] Preferably, the preset blackbody radiation reference calibration mapping relationship is pre-configured before performing temperature inversion conversion. The configuration process is as follows: First, the infrared acquisition device collects blackbody radiation response records formed by the blackbody radiation reference object under different reference temperature states. Then, blackbody grayscale mapping features with the same data format as the grayscale mapping features are extracted from the blackbody radiation response records. Subsequently, the blackbody grayscale mapping features are mapped and organized with the different reference temperature states to form the blackbody radiation reference calibration mapping relationship. The blackbody radiation response records express the grayscale level changes formed when the infrared acquisition device faces a blackbody radiation reference object with known thermal radiation states. The blackbody grayscale mapping features express the correspondence between the grayscale levels in the blackbody radiation response records and the different reference temperature states. The blackbody radiation reference calibration mapping relationship expresses the conversion relationship between grayscale level intervals and absolute temperature intervals. After being pre-configured, the blackbody radiation reference calibration mapping relationship continues to serve as the calibration reference when the grayscale mapping features undergo temperature inversion conversion, enabling the grayscale mapping features to be converted into pixel absolute temperature values ​​based on the infrared radiation response rules of the infrared acquisition device itself.

[0037] Preferably, the blackbody radiation reference calibration mapping relationship is not directly based on a single global empirical conversion relationship, but is segmented and organized according to the imaging channel state, lens field of view state, and acquisition distance state of the infrared acquisition device; wherein, the imaging channel state comes from the infrared imaging channel used by the infrared acquisition device when acquiring the blackbody radiation response record, the lens field of view state comes from the field of view coverage range corresponding to the infrared acquisition device when acquiring the blackbody radiation response record, and the acquisition distance state comes from the spatial interval between the infrared acquisition device and the blackbody radiation reference object when acquiring the blackbody radiation response record. After the imaging channel state, the lens field of view state, and the acquisition distance state are jointly written into the blackbody radiation reference calibration mapping relationship, the blackbody radiation reference calibration mapping relationship can form a correspondence with the lens field of view and the acquisition distance in the shooting pose parameters. Thus, when the temperature inversion conversion processes the grayscale mapping feature, it not only converts according to the grayscale level in the grayscale mapping feature, but also selects a matching grayscale temperature conversion relationship according to the lens field of view and the acquisition distance corresponding to the contour alignment heatmap to which the grayscale mapping feature belongs. The grayscale temperature conversion relationship continues to participate in the formation of the absolute temperature value of the pixel.

[0038] Preferably, when performing temperature inversion transformation on the grayscale mapping feature according to the preset blackbody radiation reference calibration mapping relationship, firstly, based on the contour-aligned heatmap corresponding to the grayscale mapping feature, the shooting pose parameters corresponding to the contour-aligned heatmap are read; then, based on the lens field of view and shooting distance in the shooting pose parameters, a grayscale temperature conversion relationship matching the contour-aligned heatmap is selected from the blackbody radiation reference calibration mapping relationship; subsequently, the grayscale levels in the grayscale mapping feature are matched with the grayscale temperature conversion relationship to form the absolute temperature value of the pixel. The absolute temperature value of the pixel is then written into the corresponding matrix position according to the pixel arrangement relationship, and organized according to the arrangement order of the contour-aligned heatmap in the pose normalized heatmap sequence, thereby obtaining the absolute temperature distribution matrix sequence. Each absolute temperature distribution matrix in the absolute temperature distribution matrix sequence corresponds to a contour-aligned heatmap in the pose-normalized heatmap sequence. The row and column directions in the absolute temperature distribution matrix correspond to the row and column directions of the pixels in the contour-aligned heatmap, respectively. The matrix element at the intersection of the row and column directions expresses the absolute temperature value of the corresponding pixel. The matrix element continues to participate in subsequent feature association and binding.

[0039] Preferably, after the absolute temperature distribution matrix sequence is formed, temperature consistency processing is performed on the absolute temperature distribution matrix sequence. This temperature consistency processing first identifies the absolute temperature values ​​of pixels mapped to the corresponding positions on the outer surface of the target power equipment based on their corresponding positions in different contour-aligned heatmaps. Then, based on the lens field of view, shooting distance, and imaging orientation in the shooting pose parameters, the absolute temperature values ​​of pixels mapped to the corresponding positions on the outer surface of the same target power equipment are processed to identify viewing angle differences, thus forming a consistent absolute temperature value for the corresponding position on the outer surface of the same target power equipment. This viewing angle difference processing ensures that the absolute temperature values ​​of multiple pixels at the corresponding positions on the outer surface of the same target power equipment are compared according to their corresponding lens field of view, shooting distance, and imaging orientation. The consistent absolute temperature value is then written into the matrix elements related to the corresponding positions on the outer surface of the target power equipment in the absolute temperature distribution matrix sequence, enabling the absolute temperature distribution matrix sequence to reduce temperature expression shifts caused by differences in acquisition distance and imaging orientation at different viewing angles for the corresponding positions on the outer surface of the same target power equipment. The absolute temperature distribution matrix sequence after temperature consistency processing continues to participate in the feature association binding between the absolute temperature distribution matrix sequence and the pose normalized heat map sequence.

[0040] Preferably, when binding the absolute temperature distribution matrix sequence with the pose-normalized heat map sequence for feature association, firstly, the contour-aligned heat map in the pose-normalized heat map sequence is used as the image index object to read the pixel arrangement relationship, the corresponding position on the outer surface of the target power equipment, and the infrared heat map acquisition time corresponding to the contour-aligned heat map; then, the absolute temperature distribution matrix in the absolute temperature distribution matrix sequence is used as the temperature index object to read the matrix elements in the absolute temperature distribution matrix corresponding to the pixel arrangement relationship; subsequently, the contour-aligned heat map, the corresponding position on the outer surface of the target power equipment, the infrared heat map acquisition time, and the absolute temperature values ​​of the pixels expressed by the matrix elements are bound together to form an infrared temperature spatial association record. The infrared temperature spatial correlation record records the correspondence between the contour-aligned heat map, the corresponding position on the outer surface of the target power equipment, the acquisition time of the infrared heat map, and the absolute temperature value of the pixel. The infrared temperature spatial correlation record continues to serve as a component of the multi-view infrared thermal image data source, enabling the multi-view infrared thermal image data source to simultaneously express the image source of each contour-aligned heat map, the absolute temperature value of each pixel, and the corresponding position on the outer surface of the target power equipment to which each pixel is attached.

[0041] Preferably, the feature association binding further synchronizes the absolute temperature values ​​of pixels in the absolute temperature distribution matrix sequence with the viewpoint sources in the pose-normalized thermal image sequence. Specifically, the contour-aligned thermal images in the pose-normalized thermal image sequence are first arranged sequentially according to the infrared thermal image acquisition time to form a contour-aligned thermal image sequential arrangement relationship. Then, the absolute temperature distribution matrices in the absolute temperature distribution matrix sequence are arranged accordingly according to the contour-aligned thermal image sequential arrangement relationship, thereby making each contour-aligned thermal image correspond to an absolute temperature distribution matrix with the same sequence number. This same sequence number correspondence is further bound to the corresponding position on the outer surface of the target power equipment to form a multi-view infrared thermal image data source in which the infrared thermal image acquisition time, the corresponding position on the outer surface of the target power equipment, and the absolute temperature values ​​of pixels correspond to each other. In this way, the multi-view infrared thermal image data source includes not only the absolute temperature value of the pixel, but also the viewpoint source of the absolute temperature value of the pixel and the corresponding position of the target power equipment outer surface of the pixel. When the multi-view infrared thermal image data source is decoupled using the environmental radiation elimination matrix, the absolute temperature values ​​of pixels from different viewpoints can be compared and separated based on the corresponding position of the outer surface of the same target power equipment.

[0042] Preferably, in the power distribution equipment inspection scenario, when the pose normalized heat map sequence includes a contour-aligned heat map corresponding to the outer surface of the enclosure, a contour-aligned heat map corresponding to the outer surface of the heat sink, and a contour-aligned heat map corresponding to the outer surface of the wiring end, the gray-scale mapping features corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end are first obtained from the above different contour-aligned heat maps respectively; then, according to the blackbody radiation reference calibration mapping relationship, the gray-scale mapping features corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end are respectively subjected to temperature inversion transformation to form an absolute temperature distribution matrix corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end; subsequently, the absolute temperature distribution matrix corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end are written into the absolute temperature distribution matrix sequence according to the corresponding position of the outer surface of the target power equipment. The absolute temperature distribution matrix sequence is further associated and bound with the pose-normalized thermal image sequence, enabling the multi-view infrared thermal image data source to distinguish the absolute temperature values ​​of pixels on the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end at corresponding positions on the outer surface of different target power equipment.

[0043] Preferably, the multi-view infrared thermal image data source is jointly formed by the pose-normalized thermal image sequence, the grayscale mapping feature, the absolute temperature distribution matrix sequence, the corresponding position on the outer surface of the target power equipment, the infrared thermal image acquisition time, and the infrared temperature spatial association record; wherein, the pose-normalized thermal image sequence expresses the infrared image content that has been contour-aligned, the grayscale mapping feature expresses the intensity of the infrared radiation response formed by the infrared acquisition device at each pixel, the absolute temperature distribution matrix sequence expresses the absolute temperature value of the pixel after temperature inversion transformation, the corresponding position on the outer surface of the target power equipment expresses the spatial attachment position of the absolute temperature value of the pixel in the three-dimensional structure of the target power equipment, the infrared thermal image acquisition time expresses the viewing angle acquisition sequence corresponding to the absolute temperature value of the pixel, and the infrared temperature spatial association record expresses the binding relationship between the pose-normalized thermal image sequence, the grayscale mapping feature, the absolute temperature distribution matrix sequence, the corresponding position on the outer surface of the target power equipment, and the infrared thermal image acquisition time. The multi-view infrared thermal image data source continues to serve as the basis for feature decoupling of the environmental radiation elimination matrix, enabling the subsequently generated infrared feature map of the target power equipment to be built on the data that has already been calibrated for thermal intensity and spatial correlation.

[0044] Optionally, the step of using the environmental radiation removal matrix to decouple features from the multi-view infrared thermal image data source and generate an infrared feature map of the target power equipment includes: Obtain the pre-configured environmental radiation rejection matrix, wherein the pre-configured environmental radiation rejection matrix is ​​determined by the correspondence between the background radiation characteristics of the inspection environment and the infrared imaging channel, and the background radiation characteristics of the inspection environment and the correspondence between the infrared imaging channel jointly define the distribution of matrix elements in the pre-configured environmental radiation rejection matrix; The environmental radiation removal matrix is ​​applied to the multi-view infrared thermal image data source to establish an independent component analysis matrix. The energy weight distribution ratio of different heat source components is determined by performing blind source separation iterative calculation on the multi-view infrared thermal image data source using the independent component analysis matrix. The environmental interference component is determined based on the energy weight distribution ratio, and the environmental interference component is removed from the multi-view infrared thermal image data source to generate the infrared feature map of the target power equipment.

[0045] Preferably, before obtaining the pre-configured environmental radiation elimination matrix, the background radiation characteristics of the inspection environment are first constructed based on the inspection scene corresponding to the multi-view infrared thermal imaging data source. The background radiation characteristics of the inspection environment originate from the non-equipment outer surface area adjacent to the corresponding position of the target power equipment's outer surface in the multi-view infrared thermal imaging data source, as well as the non-equipment outer surface area collected by the infrared acquisition device under the same inspection path. Specifically, infrared temperature spatial correlation records are first read from the multi-view infrared thermal imaging data source, and then the corresponding position of the target power equipment's outer surface and the non-equipment outer surface area are distinguished according to the infrared temperature spatial correlation records. Subsequently, the background infrared intensity distribution, background temperature gradual change trend, and background reflection response distribution are extracted from the non-equipment outer surface area to form the background radiation characteristics of the inspection environment. The background infrared intensity distribution is used to express the intensity of environmental radiation in the non-equipment outer surface area, the background temperature gradual change trend is used to express the continuous temperature change state in the non-equipment outer surface area, and the background reflection response distribution is used to express the infrared response state of the non-equipment outer surface area to the reflection of surrounding heat sources. The background infrared intensity distribution, the background temperature gradual change trend, and the background reflection response distribution jointly participate in the formation of the inspection environment background radiation characteristics, enabling the inspection environment background radiation characteristics to characterize the sky radiation distribution, ground reflection distribution, and surrounding structure thermal reflection distribution in the inspection scene corresponding to the multi-view infrared thermal image data source. The inspection environment background radiation characteristics further participate in the construction of the pre-configured environmental radiation elimination matrix, enabling the pre-configured environmental radiation elimination matrix to form corresponding elimination criteria for the sky radiation distribution, ground reflection distribution, and surrounding structure thermal reflection distribution in the inspection scene corresponding to the multi-view infrared thermal image data source, rather than using a uniform background subtraction method unrelated to the inspection scene corresponding to the multi-view infrared thermal image data source.

[0046] Preferably, the infrared imaging channel correspondence is pre-configured based on the channel response record formed when the infrared acquisition device images the multi-view infrared thermal image data source. The channel response record includes the correspondence between the infrared imaging channels of the infrared acquisition device and the contour-aligned heatmaps in the multi-view infrared thermal image data source, the correspondence between the infrared imaging channels and the pixel arrangement, and the correspondence between the infrared imaging channels and the absolute temperature values ​​of the pixels. Further, the grayscale response difference of each infrared imaging channel in different contour-aligned heatmaps is read according to the infrared imaging channel correspondence, and the grayscale response difference is correlated with the background radiation characteristics of the inspection environment to form a channel environment coupling record. The grayscale response difference originates from the grayscale response change formed by the same infrared imaging channel in different contour-aligned heatmaps of the corresponding position on the outer surface of the same target power equipment or in a non-equipment outer surface area. The grayscale response difference is further correlated with the background infrared intensity distribution, background temperature gradual change trend, and background reflection response distribution in the background radiation characteristics of the inspection environment, so that the channel environment coupling record can express the coupling relationship between the infrared imaging channels and the background radiation characteristics of the inspection environment. The channel environment coupling record continues to participate in the determination of the matrix element distribution of the pre-configured environmental radiation elimination matrix, so that each matrix element in the pre-configured environmental radiation elimination matrix can correspond to the coupling relationship between the infrared imaging channel, the corresponding position on the outer surface of the target power equipment, and the background radiation characteristics of the inspection environment.

[0047] Preferably, the pre-configured environmental radiation rejection matrix is ​​jointly defined by the background radiation characteristics of the inspection environment and the correspondence between the infrared imaging channels; wherein, the matrix rows of the pre-configured environmental radiation rejection matrix correspond to the contour-aligned heatmap or infrared imaging channel in the multi-view infrared thermal image data source, the matrix columns of the pre-configured environmental radiation rejection matrix correspond to the corresponding position on the outer surface of the target power equipment or the non-equipment outer surface area, and the matrix elements at the intersection of the matrix rows and columns in the pre-configured environmental radiation rejection matrix express the rejection weight of the corresponding infrared imaging channel at the corresponding position on the outer surface of the target power equipment affected by the background radiation characteristics of the inspection environment. The rejection weight is jointly determined by the background infrared intensity distribution, the background temperature gradual change trend, the background reflection response distribution, and the channel environment coupling record; wherein, the background infrared intensity distribution provides the source of the environmental radiation intensity of the rejection weight, the background temperature gradual change trend provides the source of the environmental temperature change of the rejection weight, the background reflection response distribution provides the source of the reflected radiation of the rejection weight, and the channel environment coupling record provides the source of the channel response of the rejection weight in the infrared imaging channel. The elimination weights continue to participate in subsequent feature decoupling, enabling the environmental background radiation components in the multi-view infrared thermal imaging data source to be characterized separately during the matrix operation process.

[0048] Preferably, when the pre-configured environmental radiation elimination matrix is ​​applied to the multi-view infrared thermal image data source, the pose-normalized thermal image sequence, the absolute temperature distribution matrix sequence, the corresponding position of the target power equipment's outer surface, the infrared thermal image acquisition time, and the infrared temperature spatial association record are first read from the multi-view infrared thermal image data source. Then, according to the infrared temperature spatial association record, the absolute temperature values ​​of pixels at the corresponding position of the same target power equipment's outer surface in different contour-aligned thermal images are arranged in the same position to form in-position infrared temperature arrangement data. The in-position infrared temperature arrangement data expresses the arrangement state of the absolute temperature values ​​of pixels at the corresponding position of the same target power equipment's outer surface in different infrared thermal image acquisition times and different contour-aligned thermal images. The in-position infrared temperature arrangement data continues to interact with the matrix elements in the pre-configured environmental radiation elimination matrix, so that the absolute temperature value of each pixel at the corresponding position of the target power equipment's outer surface can be decomposed into temperature components associated with the heat-generating components of the target power equipment itself, and environmental background radiation components associated with the background radiation characteristics of the inspection environment. After the above processing, environmental radiation constraint data is formed. The environmental radiation constraint data expresses the separation constraint relationship between the heat-generating components of the target power equipment and the environmental background radiation components at corresponding positions on the outer surface of the same target power equipment. The environmental radiation constraint data continues to participate in the establishment of the independent component analysis matrix.

[0049] Preferably, the independent component analysis matrix is ​​not obtained by directly decomposing the original two-dimensional infrared image, but is established based on the environmental radiation constraint data, the isotopic infrared temperature arrangement data, and the correspondence of the infrared imaging channels. The rows of the independent component analysis matrix correspond to candidate sources of different heat source components, including candidate sources of heat generation from the target power equipment itself, candidate sources of environmental background radiation, and candidate sources of external reflected radiation. The columns of the independent component analysis matrix correspond to the corresponding positions on the outer surface of the target power equipment in the multi-view infrared thermal image data source. The matrix elements at the intersection of the rows and columns express the participation intensity of the corresponding candidate sources of different heat source components at the corresponding positions on the outer surface of the target power equipment. The participation intensity originates from the absolute temperature values ​​of pixels in the isotopic infrared temperature arrangement data, the separation constraint relationships in the environmental radiation constraint data, and the channel response sources in the correspondence of the infrared imaging channels. The participation intensity continues to serve as the basis for adjusting the energy weight distribution ratio of different heat source components in the blind source separation iterative calculation. After the independent component analysis matrix is ​​established, it continues to receive the removal weights of the pre-configured environmental radiation removal matrix, so that the independent component analysis matrix can perform calculations around the separable relationship between the environmental background radiation components and the heat generation components of the target power equipment body in subsequent blind source separation iteration calculations.

[0050] Preferably, when performing blind source separation iterative calculations on the multi-view infrared thermal image data source using the independent component analysis matrix, the isotopic infrared temperature arrangement data is first input into the independent component analysis matrix, allowing the independent component analysis matrix to initially separate the isotopic infrared temperature arrangement data according to the candidate sources of different heat source components, forming initial heat source component separation data. Then, according to the removal weights in the pre-configured environmental radiation removal matrix, the environmental background radiation components in the initial heat source component separation data corresponding to the background radiation characteristics of the inspection environment are iteratively subtracted, and the heat-generating components of the target power equipment body corresponding to the continuous distribution at the corresponding position on the outer surface of the target power equipment are retained and organized to form iterative heat source component separation data. The iterative heat source component separation data continues to be compared with the infrared temperature spatial association record, so that the iterative heat source component separation data after each iteration still maintains the spatial correspondence with the corresponding position on the outer surface of the target power equipment; the heat-generating components of the target power equipment body in the iterative heat source component separation data continue to participate in the generation of the infrared feature map of the target power equipment, and the environmental background radiation components in the iterative heat source component separation data continue to participate in the determination of environmental interference components.

[0051] Preferably, in the blind source separation iterative operation, the energy weight distribution ratio of different heat source components is determined based on the participation intensity of the heat source component iterative separation data at corresponding positions on the outer surface of different target power equipment; wherein, the energy weight distribution ratio of different heat source components includes the energy weight distribution ratio corresponding to the heat-generating components of the target power equipment body, the energy weight distribution ratio corresponding to the environmental background radiation components, and the energy weight distribution ratio corresponding to the external reflected radiation components. Specifically, the participation intensity corresponding to the candidate sources of each type of different heat source component in the heat source component iterative separation data is first read, and then the participation intensity is normalized according to the corresponding position on the outer surface of the target power equipment to determine the energy proportion relationship of different heat source components at the corresponding position on the outer surface of the same target power equipment, thereby forming the energy weight distribution ratio of different heat source components. The energy proportion relationship expresses the relative contribution of different heat source components at corresponding positions on the outer surface of the same target power equipment. The energy proportion relationship continues to participate in the formation of the energy weight distribution ratio of the different heat source components. The energy weight distribution ratio of the different heat source components continues to participate in the determination of the environmental interference component, so that the determination of the environmental interference component no longer depends solely on the local brightness difference in the two-dimensional image, but on the energy proportion relationship of different heat source components at corresponding positions on the outer surface of the same target power equipment.

[0052] Preferably, when determining the environmental interference component based on the energy weight distribution ratio of the different heat source components, the energy weight distribution ratio of the different heat source components is first compared with the background radiation characteristics of the inspection environment to screen out the heat source components that are continuously present in the non-equipment outer surface area and change synchronously with the viewing angle at the corresponding position on the outer surface of the target power equipment, as candidate components for environmental interference; then, the correspondence of the infrared imaging channels is used to determine whether the candidate components for environmental interference are concentrated in the infrared imaging channels that are highly affected by the background radiation characteristics of the inspection environment; when the candidate components for environmental interference simultaneously satisfy the conditions of corresponding to the background radiation characteristics of the inspection environment, corresponding to the infrared imaging channels that are highly affected by the background radiation characteristics of the inspection environment, and not having a stable spatial attachment relationship with the structural spatial coordinates in the three-dimensional structure of the target power equipment, the candidate components for environmental interference are determined as the environmental interference component. The environmental interference component includes the portion of the ambient background radiation component and the external reflected radiation component that does not belong to the heat generation component of the target power equipment body. The environmental interference component continues to be the object to be removed from the multi-view infrared thermal image data source, so that the infrared feature map of the target power equipment generated subsequently can reduce the interference of the ambient background radiation component and the external reflected radiation component on the heat generation component of the target power equipment body.

[0053] Preferably, when removing the environmental interference component from the multi-view infrared thermal image data source, the distribution position of the environmental interference component on the corresponding position of each target power device's outer surface is first determined according to the infrared temperature spatial correlation record. Then, the removal intensity of the environmental interference component at the corresponding distribution position is determined according to the removal weight in the pre-configured environmental radiation removal matrix. Subsequently, the absolute temperature values ​​of pixels in the multi-view infrared thermal image data source are subtracted for the environmental interference component according to the removal intensity to form the target power device's body temperature characterization data. The target power device's body temperature characterization data expresses the absolute temperature values ​​of pixels that still maintain a spatial correspondence with the corresponding position on the target power device's outer surface after subtracting the environmental interference component. The target power device's body temperature characterization data is further spatially rearranged according to the corresponding position on the target power device's outer surface, so that the target power device's body temperature characterization data maintains a correspondence with the structural spatial coordinates in the target power device's three-dimensional structure, thereby generating the target power device's infrared feature map. The infrared feature map of the target power equipment represents the infrared distribution state of the target power equipment body after removing the environmental interference component. The infrared feature map of the target power equipment continues to participate in the generation of the subsequent surface temperature mapping feature set of the target power equipment.

[0054] Preferably, in the power distribution equipment inspection scenario, when the multi-view infrared thermal image data source includes the absolute temperature values ​​of pixels corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals, the background radiation features of the inspection environment are extracted from the non-equipment outer surface areas surrounding the outer surface of the enclosure, the non-equipment outer surface areas outside the gaps on the outer surface of the heat sink, and the non-equipment outer surface areas surrounding the outer surface of the wiring terminals. The pre-configured environmental radiation elimination matrix forms elimination weights based on the background infrared intensity distribution, background temperature gradual change trend, and background reflection response distribution of the aforementioned different non-equipment outer surface areas; the independent component analysis matrix then performs blind source separation iterative calculations on the absolute temperature values ​​of pixels on the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals according to the elimination weights to determine the energy weight distribution ratio of different heat source components. After determining the environmental interference component based on the energy weight distribution ratio of the different heat source components, the environmental interference component is eliminated from the multi-view infrared thermal image data source to generate infrared feature maps of the target power equipment that can respectively express the heat generation distribution of the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals.

[0055] Optionally, the step of generating a surface temperature mapping feature set of the target power equipment based on the infrared feature map of the target power equipment includes: Determine the thermal gradient features in the infrared feature map of the target power equipment; Obtain a preset temperature threshold that matches the device type and operating conditions of the target power equipment, wherein the preset temperature threshold is pre-configured based on the device type and operating conditions of the target power equipment; Spatial connectivity analysis is performed on the thermal gradient features to delineate target temperature connectivity regions with temperature values ​​greater than the preset temperature threshold. The contour boundary coordinates corresponding to the target temperature connected region are determined, and the contour boundary coordinates are structurally recombined with the infrared feature map of the target power equipment to generate the surface temperature mapping feature set of the target power equipment.

[0056] Preferably, when generating the surface temperature mapping feature set of the target power equipment based on the infrared feature map of the target power equipment, the corresponding position of the outer surface of the target power equipment, the absolute temperature value of the pixel, and the body temperature characterization data of the target power equipment are first read from the infrared feature map of the target power equipment. The infrared feature map of the target power equipment originates from the spatial rearrangement result after removing environmental interference components from the multi-view infrared thermal image data source. The corresponding position of the outer surface of the target power equipment represents the spatial attachment position of the absolute temperature value of the pixel in the three-dimensional structure of the target power equipment. The body temperature characterization data of the target power equipment represents the absolute temperature value of the pixel that still maintains a corresponding relationship with the corresponding position of the outer surface of the target power equipment after removing environmental interference components. Based on the correspondence between the corresponding positions on the outer surface of the target power equipment and the absolute temperature values ​​of the pixels, the temperature changes between corresponding positions on the outer surfaces of adjacent target power equipment in the infrared feature map of the target power equipment are read to determine the thermal gradient features in the infrared feature map of the target power equipment. The thermal gradient features are used to express the distribution of temperature change amplitude, distribution of temperature change direction, and continuous temperature change relationship between different spatial positions on the outer surface of the target power equipment. The distribution of temperature change amplitude, distribution of temperature change direction, and continuous temperature change relationship continue to participate in subsequent spatial connectivity analysis.

[0057] Preferably, the process of determining the thermal gradient features is not merely a brightness difference analysis of adjacent pixels in a two-dimensional image. Instead, it uses the structural spatial coordinates in the three-dimensional structure of the target power equipment as a reference. First, it establishes an adjacency relationship between the outer surfaces of the target power equipment based on their corresponding positions. This adjacency relationship expresses the spatial adjacency between corresponding positions on the outer surfaces of adjacent target power equipment on the outer surfaces of the enclosure, the heat sink, and the wiring terminals. Then, it reads the absolute temperature values ​​of pixels at corresponding positions on the outer surfaces of adjacent target power equipment based on the adjacency relationship, and performs differential processing on these values ​​to form a temperature variation amplitude distribution. Subsequently, it forms a temperature variation direction distribution based on the extension direction of the temperature variation amplitude distribution within the adjacency relationship of the target power equipment. Finally, it forms a continuous temperature variation relationship based on the continuous state of the temperature variation amplitude distribution and the temperature variation direction distribution between corresponding positions on the outer surfaces of adjacent target power equipment. The temperature change amplitude distribution, the temperature change direction distribution, and the temperature continuous change relationship together constitute the thermal gradient feature, which enables the thermal gradient feature to reflect the spatial expansion trend of heat on the outer surface of the target power equipment, rather than just reflecting the local grayscale changes in the two-dimensional image.

[0058] Preferably, before obtaining a preset temperature threshold matching the device type and operating condition of the target power equipment, a preset temperature threshold configuration record is first established. This preset temperature threshold configuration record includes the device type of the target power equipment, the operating condition of the target power equipment, the structural region to which the outer surface of the target power equipment belongs, and temperature judgment boundaries corresponding to the device type, the operating condition, and the structural region to which the outer surface of the target power equipment belongs. The device type is used to distinguish the temperature distribution differences of enclosure-type structures, heat sink-type structures, and terminal-type structures under normal heating conditions. The operating condition is used to distinguish the surface temperature changes of the target power equipment under different load conditions, heat dissipation conditions, and inspection periods. The structural region to which the outer surface of the target power equipment belongs expresses the outer surface of the enclosure, the outer surface of the heat sink, or the outer surface of the terminal to which the corresponding location on the outer surface of the target power equipment belongs. The temperature judgment boundary is used to limit the temperature conditions that can be considered as candidate locations for abnormal heating under the given device type, operating condition, and structural region to which the outer surface of the target power equipment belongs. By configuring the preset temperature threshold, a preset temperature threshold can be provided for the absolute temperature value of the pixel in the infrared feature map of the target power equipment before performing spatial connectivity analysis. This preset temperature threshold matches the equipment type, operating conditions, and the structural region to which the outer surface of the target power equipment belongs.

[0059] Preferably, the process of obtaining the preset temperature threshold is as follows: First, the structural region corresponding to the location on the outer surface of the target power equipment is read from the three-dimensional structure of the target power equipment. Then, the corresponding structural thermal response attribute is determined according to the equipment type of the target power equipment and the structural region on the outer surface of the target power equipment. Combined with the operating conditions of the target power equipment, a preset temperature threshold corresponding to the structural thermal response attribute and the operating conditions is selected from the preset temperature threshold configuration record. The structural thermal response attribute is used to express the response differences of different structural regions on the outer surface of the target power equipment during heating, heat dissipation, and heat transfer. The structural thermal response attribute continues to participate in the selection of the preset temperature threshold, so that the preset temperature threshold is correspondingly saved with the location on the outer surface of the target power equipment. When performing spatial connectivity analysis on the thermal gradient features subsequently, the absolute temperature value of a pixel at the corresponding location on the outer surface of the same target power equipment can be compared with the preset temperature threshold corresponding to the structural region on the outer surface of the target power equipment at that location. Unlike two-dimensional image processing methods that use a globally uniform threshold, the preset temperature threshold establishes a correspondence between the target power equipment type, the target power equipment operating condition, the structural region to which the target power equipment's outer surface belongs, and the corresponding position on the target power equipment's outer surface. This allows the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals to participate in the delineation of the target temperature connectivity region according to their respective structural thermal response attributes.

[0060] Preferably, when performing spatial connectivity analysis on the thermal gradient features, the absolute temperature value of each pixel corresponding to the outer surface of the target power equipment in the infrared feature image of the target power equipment is first compared with a preset temperature threshold corresponding to that location on the outer surface of the target power equipment. When the absolute temperature value of the pixel is greater than the preset temperature threshold, the location corresponding to the outer surface of the target power equipment is marked as a candidate high-temperature outer surface location. The candidate high-temperature outer surface locations are then further analyzed in relation to the temperature change amplitude distribution, temperature change direction distribution, and temperature continuous change relationship in the thermal gradient features to determine whether multiple candidate high-temperature outer surface locations have spatial adjacency and temperature continuous change relationships within the three-dimensional structure of the target power equipment. Thus, the spatial connectivity analysis is not a simple connection of isolated bright pixels in the two-dimensional thermal image, but rather a joint judgment of the spatial adjacency, temperature change direction distribution, and temperature continuous change relationship between candidate high-temperature outer surface locations based on the location corresponding to the outer surface of the target power equipment, the candidate high-temperature outer surface locations, and the thermal gradient features.

[0061] Preferably, when forming the target temperature connected region, firstly, based on the adjacency relationship of the target power equipment's outer surface, candidate high-temperature outer surface positions with spatial adjacency are grouped into the same candidate connected region; the candidate connected region originates from the connection state of the candidate high-temperature outer surface positions in the adjacency relationship of the target power equipment's outer surface, and the candidate connected region continues to be screened by the thermal gradient features. Then, based on the temperature change direction distribution in the thermal gradient features, it is determined whether the candidate high-temperature outer surface positions in the same candidate connected region are continuously distributed along similar temperature expansion directions; subsequently, based on the temperature change amplitude distribution and continuous temperature change relationship in the thermal gradient features, local temperature abrupt change points formed by a single candidate high-temperature outer surface position or discrete candidate high-temperature outer surface positions are identified, and the local temperature abrupt change points are excluded from the candidate connected region to delineate the target temperature connected region with a temperature value greater than the preset temperature threshold. The local temperature abrupt change points originate from candidate high-temperature outer surface locations that do not satisfy spatial adjacency or continuous temperature change relationships. After the local temperature abrupt change points are excluded, the target temperature connected region can express the heat-generating areas on the outer surface of the target power equipment that meet the preset temperature threshold conditions and simultaneously have spatial adjacency and continuous temperature change relationships. The target temperature connected region continues to participate in the determination of the contour boundary coordinates.

[0062] Preferably, when determining the contour boundary coordinates corresponding to the target temperature connected region, firstly, along the outer periphery of the target temperature connected region, the positions of candidate high-temperature outer surfaces located inside the target temperature connected region are searched, as well as the corresponding positions of the target power equipment outer surfaces adjacent to the candidate high-temperature outer surfaces but not belonging to the target temperature connected region; then, the boundary position between the above adjacent positions is taken as the boundary position of the target temperature connected region, and the contour boundary coordinates are determined according to the structural spatial coordinates of the boundary position of the target temperature connected region in the three-dimensional structure of the target power equipment. The boundary position of the target temperature connected region originates from the spatial boundary between the target temperature connected region and non-target temperature connected regions, and the boundary position of the target temperature connected region continues to participate in the formation of the contour boundary coordinates; the contour boundary coordinates are used to express the spatial boundary of the target temperature connected region in the three-dimensional structure of the target power equipment, and the contour boundary coordinates are further structurally recombined with the infrared feature map of the target power equipment, so that the target temperature connected region is no longer just a planar boundary in a two-dimensional image, but a spatial boundary that can correspond to the outer surface structure of the target power equipment.

[0063] Preferably, the contour boundary coordinates include the boundary coordinate arrangement relationship, the structural region to which the boundary coordinates belong, and the absolute temperature value of the pixel corresponding to the boundary coordinates. The boundary coordinate arrangement relationship expresses the order in which multiple contour boundary coordinates are arranged along the outer periphery of the target temperature connected region. The structural region to which the boundary coordinates belong expresses the structural region where the contour boundary coordinates are located on the outer surface of the housing, the outer surface of the heat sink, or the outer surface of the wiring terminal. The absolute temperature value of the pixel corresponding to the boundary coordinates originates from the position on the outer surface of the target power equipment corresponding to the contour boundary coordinates in the infrared feature image of the target power equipment. The boundary coordinate arrangement relationship, the structural region to which the boundary coordinates belong, and the absolute temperature value of the pixel corresponding to the boundary coordinates all participate in subsequent structured reconstruction. The boundary coordinate arrangement relationship is used to maintain the order of the outer periphery of the target temperature connected region. The structural region to which the boundary coordinates belong is used to distinguish the structural region on the outer surface of the target power equipment that the target temperature connected region crosses. The absolute temperature value of the pixel corresponding to the boundary coordinates is used to express the temperature distribution on the outer periphery of the target temperature connected region. Through the joint participation of the above contents in subsequent structured reconstruction, the surface temperature mapping feature set of the target power equipment can simultaneously describe the boundary morphology, spatial location, and temperature distribution of the heat-generating region.

[0064] Preferably, when structurally recombining the contour boundary coordinates with the infrared feature image of the target power equipment, the following steps are taken: First, using the contour boundary coordinates as boundary indexes, the absolute temperature values ​​of pixels within the target temperature connected region, the corresponding positions on the outer surface of the target power equipment, and the thermal gradient features are read from the infrared feature image of the target power equipment. Then, the absolute temperature values ​​of pixels within the target temperature connected region are spatially arranged according to their corresponding positions on the outer surface of the target power equipment, and the thermal gradient features are directionally associated according to the boundary coordinate arrangement relationship to form a temperature structure record for the target temperature connected region. This temperature structure record includes the correspondence between the absolute temperature values ​​of pixels within the target temperature connected region, the corresponding positions on the outer surface of the target power equipment within the target temperature connected region, the thermal gradient features within the target temperature connected region, and the contour boundary coordinates. The temperature structure record is then further merged and organized with the contour boundary coordinates so that the surface temperature mapping feature set of the target power equipment can express the correspondence between the internal temperature distribution, the outer boundary position, and the thermal gradient extension direction of the target temperature connected region.

[0065] Preferably, the target power equipment surface temperature mapping feature set is jointly formed by the target power equipment infrared feature map, the thermal gradient feature, the preset temperature threshold, the target temperature connected region, the contour boundary coordinates, and the temperature structure record of the target temperature connected region. Specifically, the target power equipment infrared feature map provides the infrared distribution state of the target power equipment body after removing environmental interference components; the thermal gradient feature provides the direction and magnitude of temperature change on the outer surface of the target power equipment; the preset temperature threshold provides the temperature determination boundary corresponding to the equipment type, operating condition, and structural region of the target power equipment's outer surface; the target temperature connected region provides heating areas that meet the temperature conditions and have spatial connectivity; the contour boundary coordinates provide the spatial boundary of the target temperature connected region within the three-dimensional structure of the target power equipment; and the temperature structure record of the target temperature connected region provides the correlation between the internal temperature distribution and the position of the outer boundary. The target power equipment surface temperature mapping feature set continues to participate in subsequent structural fusion processing with the three-dimensional structure of the target power equipment, enabling the subsequent determination of the target power equipment's thermal boundary conditions based on the actual temperature distribution and spatial boundary relationships of the target power equipment's outer surface.

[0066] Preferably, in the scenario of power distribution equipment inspection, when the infrared feature map of the target power equipment includes the corresponding positions of the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end, the thermal gradient features corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end are determined respectively. Then, according to the equipment type and operating conditions of the target power equipment, the preset temperature thresholds corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end are read from the preset temperature threshold configuration record. Subsequently, spatial connectivity analysis is performed on the thermal gradient features corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end to delineate the target temperature connectivity regions corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end. By further determining the contour boundary coordinates of the target temperature connected region corresponding to the outer surface of the enclosure, the contour boundary coordinates of the target temperature connected region corresponding to the outer surface of the heat sink, and the contour boundary coordinates of the target temperature connected region corresponding to the outer surface of the wiring end, and by structurally recombining the contour boundary coordinates with the infrared feature map of the target power equipment, a target power equipment surface temperature mapping feature set that can distinguish the heat distribution of the structural region to which the outer surface of different target power equipment belongs can be generated.

[0067] Optionally, the step of structurally fusing the surface temperature mapping feature set of the target power equipment with the three-dimensional structure of the target power equipment to determine the thermal boundary conditions of the target power equipment includes: Obtain a preset physical depth level pre-configured according to the entity structure layer relationship in the three-dimensional structure of the target power equipment, wherein the entity structure layer relationship is used to define the layer position of the preset physical depth level in the three-dimensional structure of the target power equipment; The surface temperature mapping feature set of the target power equipment is hierarchically parsed according to the preset physical depth level in order to extract the surface heat dissipation gradient grid representing different preset physical depth levels. The surface heat dissipation gradient grid is mapped to the three-dimensional structure of the target power equipment in spatial coordinates to determine the three-dimensional spatial temperature distribution profile. The thermal boundary conditions of the target power equipment are determined based on the three-dimensional spatial temperature distribution profile.

[0068] Preferably, before structurally fusing the surface temperature mapping feature set of the target power equipment with the three-dimensional structure of the target power equipment, the entity structural layer relationship is first read from the three-dimensional structure of the target power equipment. This entity structural layer relationship originates from the structural spatial coordinates and material thermal conductivity properties within the three-dimensional structure of the target power equipment, and is used to express the hierarchical arrangement, spatial adjacency, and material thermal conductivity differences between the outer surface, adjacent heat transfer structural layers, and internal heat conduction structural layers of the target power equipment. The entity structural layer relationship is not an isolated equipment structure name, but rather a hierarchical description formed by correspondingly organizing the positions on the outer surface of the target power equipment, the structural spatial coordinates within the three-dimensional structure of the target power equipment, and the material thermal conductivity properties. This hierarchical description continues to participate in the configuration of the preset physical depth level as the expression content of the entity structural layer relationship, enabling the preset physical depth level to be set along the internal heat transfer direction from the outer surface of the target power equipment to its interior.

[0069] Preferably, the preset physical depth level is pre-configured based on the entity structure layer relationship. Specifically, firstly, the structural region to which the target power equipment's outer surface belongs is determined according to the entity structure layer relationship. Then, based on the adjacent heat transfer structural layers and internal heat conduction structural layers within the three-dimensional structure of the target power equipment, a preset physical depth level corresponding to the structural region to which the target power equipment's outer surface belongs is configured. The preset physical depth level is used to express the heat transfer depth position that can be resolved hierarchically when extending from the outer surface of the target power equipment to the interior of the target power equipment. The preset physical depth level includes the surface heat transfer depth near the outer surface of the target power equipment, the transition heat transfer depth in the adjacent heat transfer structural layers, and the internal heat transfer depth in the internal heat conduction structural layers. The surface heat transfer depth, the transition heat transfer depth, and the internal heat transfer depth together define the hierarchical position in the internal heat transfer direction. The preset physical depth level continues to be read in correspondence with the surface temperature mapping feature set of the target power equipment, so that subsequent hierarchical resolution can be expanded according to the structural hierarchical description defined by the entity structure layer relationship.

[0070] Preferably, when performing hierarchical parsing of the surface temperature mapping feature set of the target power equipment according to the preset physical depth level, the target temperature connected region, contour boundary coordinates, thermal gradient features, and temperature structure record of the target temperature connected region in the surface temperature mapping feature set of the target power equipment are read first. The target temperature connected region is used to define the range of the heat-generating area on the outer surface of the target power equipment; the contour boundary coordinates are used to define the spatial boundary of the target temperature connected region in the three-dimensional structure of the target power equipment; the thermal gradient features are used to express the distribution of temperature change direction and temperature change amplitude within the target temperature connected region; and the temperature structure record of the target temperature connected region is used to express the correlation between the internal temperature distribution of the target temperature connected region and the position of the outer boundary. Subsequently, the target temperature connected region, the contour boundary coordinates, the thermal gradient features, and the temperature structure record of the target temperature connected region are respectively mapped to the preset physical depth level to form a hierarchical temperature parsing record. The hierarchical temperature analysis record expresses the temperature distribution relationship of the target temperature connected region after unfolding along the preset physical depth level. The hierarchical temperature analysis record continues to participate in the extraction of the surface heat dissipation gradient grid, so that the surface temperature expression in the surface temperature mapping feature set of the target power equipment can be hierarchically described along the preset physical depth level in the direction of internal heat transfer.

[0071] Preferably, the surface heat dissipation gradient grid is obtained by meshing the hierarchical temperature analysis records. Each grid cell in the surface heat dissipation gradient grid corresponds to a position on the outer surface of a target power device and its associated preset physical depth level. Each grid cell records the absolute temperature value of a pixel at the corresponding position on the outer surface of the target power device, thermal gradient features, contour boundary coordinates, material thermal conductivity properties, and the preset physical depth level. Further, based on the temperature change direction distribution in the thermal gradient features, the temperature change direction distribution between adjacent grid cells is written into the surface heat dissipation gradient grid; based on the temperature change amplitude distribution in the thermal gradient features, the temperature change amplitude distribution between adjacent grid cells is written into the surface heat dissipation gradient grid; and based on the material thermal conductivity properties, the material thermal conductivity differences between different preset physical depth levels are written into the surface heat dissipation gradient grid. Thus, the surface heat dissipation gradient grid can characterize the spatial trend of heat dissipation from the outer surface of the target power device to adjacent heat transfer structure layers and internal heat conduction structure layers at different preset physical depth levels. The surface heat dissipation gradient grid continues to serve as input for spatial coordinate mapping processing.

[0072] Preferably, the surface heat dissipation gradient grid is not a simple copy of the surface temperature mapping feature set of the target power equipment, but rather a fusion of the surface temperature expression in the surface temperature mapping feature set of the target power equipment with the relationship between the solid structural layers in the three-dimensional structure of the target power equipment. Specifically, the outer perimeter of the target temperature connected region in the three-dimensional structure of the target power equipment is first determined based on the contour boundary coordinates. Then, the internal heat transfer direction corresponding to the outer perimeter is determined based on the preset physical depth level. Subsequently, the thermal gradient features are hierarchically arranged along the internal heat transfer direction to form the surface heat dissipation gradient grid. The grid cells in the surface heat dissipation gradient grid retain both the outer surface temperature distribution of the target temperature connected region and the correspondence between the preset physical depth level and the thermal conductivity properties of the material. This allows the surface heat dissipation gradient grid to serve as a direct input for subsequent spatial coordinate mapping processing, enabling the spatial coordinate mapping processing to simultaneously read the temperature content, structural content, and material content in the grid cells.

[0073] Preferably, when performing spatial coordinate mapping processing between the surface heat dissipation gradient grid and the three-dimensional structure of the target power equipment, the following steps are taken: First, the corresponding position, preset physical depth level, and contour boundary coordinates of the target power equipment's outer surface corresponding to each grid cell in the surface heat dissipation gradient grid are read; then, the structural spatial coordinates corresponding to the position, preset physical depth level, and contour boundary coordinates of the target power equipment's outer surface are read from the three-dimensional structure of the target power equipment; subsequently, the absolute temperature value, thermal gradient characteristics, and material thermal conductivity properties of the pixels in the grid cells are mapped to the corresponding structural spatial coordinates to form spatial temperature mapping points. These spatial temperature mapping points represent the absolute temperature value, thermal gradient characteristics, and material thermal conductivity properties of a pixel at a structural spatial coordinate position within the three-dimensional structure of the target power equipment. These spatial temperature mapping points continue to participate in the formation of spatial temperature propagation records and, through these records, participate in the determination of the three-dimensional spatial temperature distribution contour.

[0074] Preferably, the spatial coordinate mapping process further organizes the spatial adjacency relationships between the spatial temperature mapping points. Specifically, firstly, based on the structural spatial coordinates in the three-dimensional structure of the target power equipment, the spatial adjacency relationships between adjacent spatial temperature mapping points are determined. Then, based on the temperature change direction distribution and temperature change amplitude distribution in the surface heat dissipation gradient grid, the temperature propagation relationship between adjacent spatial temperature mapping points is determined. Subsequently, the temperature propagation relationship is hierarchically organized according to the material thermal conductivity properties to form a spatial temperature propagation record. The hierarchical organization of material thermal conductivity properties is used to write the material thermal conductivity differences between the outer surface of the target power equipment, adjacent heat transfer structural layers, and internal heat transfer structural layers into the temperature propagation relationship. The spatial temperature propagation record expresses the temperature transfer direction, temperature change amplitude distribution, and material thermal conductivity differences between adjacent spatial temperature mapping points. The spatial temperature propagation record continues to participate in the organization of the three-dimensional spatial temperature distribution contour, enabling the three-dimensional spatial temperature distribution contour to simultaneously reflect the structural spatial coordinates and temperature propagation relationship.

[0075] Preferably, the three-dimensional spatial temperature distribution contour is determined based on the spatial temperature mapping points and the spatial temperature propagation record. Specifically, spatial temperature mapping points with similar absolute temperature values ​​and spatial adjacency are first connected to form an initial temperature spatial contour. This initial temperature spatial contour originates from the proximity of absolute temperature values ​​and spatial adjacency between the spatial temperature mapping points. The initial temperature spatial contour is further processed by the hierarchical arrangement of the spatial temperature propagation record. Then, based on the temperature transfer direction and material thermal conductivity differences in the spatial temperature propagation record, the initial temperature spatial contour is further processed hierarchically to form a three-dimensional spatial temperature distribution contour that can express the temperature distribution relationship between the outer surface, adjacent heat transfer structural layers, and internal heat conduction structural layers of the target power equipment. The three-dimensional spatial temperature distribution contour includes the outer surface temperature contour corresponding to the target temperature connectivity region, the hierarchical temperature contour extending along a preset physical depth level, and the temperature decay contour corresponding to the material thermal conductivity properties. The outer surface temperature contour, the hierarchical temperature contour, and the temperature decay contour jointly participate in the determination of the thermal boundary conditions of the target power equipment.

[0076] Preferably, when determining the thermal boundary conditions of the target power equipment based on the three-dimensional spatial temperature distribution profile, the outer surface temperature profile, hierarchical temperature profile, and temperature decay profile in the three-dimensional spatial temperature distribution profile are first read. Then, the surface temperature boundary between the outer surface of the target power equipment and the external environment is determined based on the outer surface temperature profile. The interlayer heat transfer boundary between the outer surface of the target power equipment and adjacent heat transfer structural layers is determined based on the hierarchical temperature profile. The internal heat transfer boundary between adjacent heat transfer structural layers and internal heat conduction structural layers is determined based on the temperature decay profile and the material's thermal conductivity properties. The surface temperature boundary, the interlayer heat transfer boundary, and the internal heat transfer boundary together form the thermal boundary conditions of the target power equipment. Specifically, the surface temperature boundary represents the temperature initiation position at the outer surface of the target power equipment, the interlayer heat transfer boundary represents the hierarchical boundary position when the outer surface of the target power equipment transfers heat to adjacent heat transfer structural layers, and the internal heat transfer boundary represents the internal boundary position when adjacent heat transfer structural layers transfer heat to internal heat conduction structural layers. Therefore, the thermal boundary conditions of the target power equipment can express the temperature initiation position, temperature transfer direction, and temperature decay state of the target power equipment in the three-dimensional structural space.

[0077] Preferably, the thermal boundary condition of the target power equipment is not a single surface temperature value, but is jointly formed by the target temperature connectivity region, the contour boundary coordinates, the surface heat dissipation gradient grid, the three-dimensional spatial temperature distribution contour, and the material thermal conductivity properties. Specifically, the target temperature connectivity region provides the range of the heating area; the contour boundary coordinates provide the spatial boundary of the heating area within the three-dimensional structure of the target power equipment; the surface heat dissipation gradient grid provides the temperature dissipation trend along a preset physical depth level; the three-dimensional spatial temperature distribution contour provides the outer surface temperature contour, the layer temperature contour, and the temperature decay contour; and the material thermal conductivity properties provide the material thermal conductivity differences between the outer surface of the target power equipment, adjacent heat transfer structural layers, and internal heat conduction structural layers. All of the above components contribute to the formation of the thermal boundary condition of the target power equipment, enabling it to continue serving as input for subsequent heat conduction simulations.

[0078] Preferably, in the scenario of power distribution equipment inspection, when the surface temperature mapping feature set of the target power equipment includes the target temperature connected region corresponding to the outer surface of the enclosure, the target temperature connected region corresponding to the outer surface of the heat sink, and the target temperature connected region corresponding to the outer surface of the wiring end, firstly, according to the solid structure layer relationship in the three-dimensional structure of the target power equipment, the corresponding preset physical depth levels are configured for the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end; then, according to the corresponding preset physical depth levels, the target temperature connected regions corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end are respectively subjected to hierarchical analysis to form the corresponding surface heat dissipation gradient grid; subsequently, the corresponding surface heat dissipation gradient grid is mapped to the structural space coordinates of the three-dimensional structure of the target power equipment to form the three-dimensional spatial temperature distribution contour corresponding to the outer surface of the enclosure, the three-dimensional spatial temperature distribution contour corresponding to the outer surface of the heat sink, and the three-dimensional spatial temperature distribution contour corresponding to the outer surface of the wiring end. Based on the three-dimensional temperature distribution profiles corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals, the thermal boundary conditions of the target power equipment corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals are determined respectively, so that the different heat transfer structure relationships of the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals can participate in the subsequent heat conduction deduction.

[0079] Optionally, the step of constructing a three-dimensional temperature field distribution map of the target power equipment by performing heat conduction deduction based on the thermal boundary conditions of the target power equipment includes: The set of unsteady-state heat conduction inference rules is called to match the material thermal conductivity properties in the three-dimensional structure of the target power equipment and the thermal boundary conditions of the target power equipment. The set of unsteady-state heat conduction inference rules includes rules for heat transfer along the material thermal conductivity direction, rules for heat diffusion at the junction of structural layers, and rules for determining the attenuation of heat with spatial distance. Using the unsteady-state heat conduction deduction rule set, reverse heat conduction deduction is performed based on the thermal boundary conditions of the target power equipment to reconstruct the internal heat conduction path; Temperature field reconstruction is performed along the internal heat conduction path to extract the geometric features of the three-dimensional isothermal surface; Based on the geometric features of the three-dimensional isothermal surface, a three-dimensional spatial structure is encapsulated to construct a three-dimensional temperature field distribution map of the target power equipment.

[0080] Preferably, before performing heat conduction simulation based on the thermal boundary conditions of the target power equipment, the thermal boundary conditions of the target power equipment, the thermal conductivity properties of the materials in the three-dimensional structure of the target power equipment, and the three-dimensional spatial temperature distribution profile are read first. The thermal boundary conditions of the target power equipment include surface temperature boundaries, interlayer heat transfer boundaries, and internal heat transfer boundaries. The thermal conductivity properties of the materials are used to express the differences in thermal conductivity between the outer surface of the target power equipment, adjacent heat transfer structural layers, and internal heat conduction structural layers during heat transfer. The three-dimensional spatial temperature distribution profile is used to express the temperature distribution relationship between the outer surface of the target power equipment, adjacent heat transfer structural layers, and internal heat conduction structural layers. After correspondingly reading the thermal boundary conditions of the target power equipment, the thermal conductivity properties of the materials, and the three-dimensional spatial temperature distribution profile, a heat conduction simulation call basis is formed. This heat conduction simulation call basis is used to express the material thermal conduction direction, structural layer boundary position, and structural spatial coordinate distance change relationship that must be followed when heat travels back from the outer surface of the target power equipment to adjacent heat transfer structural layers and internal heat conduction structural layers. This heat conduction simulation call basis continues to participate in the invocation of the unsteady-state heat conduction simulation rule set.

[0081] Preferably, the unsteady-state heat conduction deduction rule set is pre-configured according to the heat conduction deduction call criteria. Specifically, firstly, the thermal conduction directions of the materials corresponding to the outer surface, adjacent heat transfer structural layers, and internal heat conduction structural layers of the target power equipment are determined based on the material's thermal conductivity properties. Then, based on the surface temperature boundary, interlayer heat transfer boundary, and internal heat transfer boundary in the thermal boundary conditions of the target power equipment, the starting position, boundary position, and ending position of heat transfer between different structural layers are determined. Subsequently, the material thermal conduction directions, the starting position, the boundary position, and the ending position are organized accordingly to form the unsteady-state heat conduction deduction rule set. The unsteady-state heat conduction deduction rule set is used to limit the order of heat transfer in the outer surface, adjacent heat transfer structural layers, and internal heat conduction structural layers of the target power equipment. The unsteady-state heat conduction deduction rule set continues to participate in the reverse heat conduction deduction.

[0082] Preferably, the unsteady-state heat conduction deduction rule set includes rules for heat transfer along the material's thermal conductivity direction, rules for heat diffusion at structural layer boundaries, and rules for determining heat attenuation with spatial distance. Specifically, the rules for heat transfer along the material's thermal conductivity direction express the directional relationship when heat is transferred within the same structural layer along the material's thermal conductivity direction; the rules for heat diffusion at structural layer boundaries express the diffusion relationship when heat enters an adjacent heat transfer structural layer from the outer surface of the target power equipment and then enters an inner heat conduction structural layer from the adjacent heat transfer structural layer; and the rules for determining heat attenuation with spatial distance express the attenuation relationship formed by heat changing with the spatial coordinate distance after leaving the surface temperature boundary. The rules for heat transfer along the material's thermal conductivity direction, the rules for heat diffusion at structural layer boundaries, and the rules for determining heat attenuation with spatial distance jointly constrain subsequent reverse heat conduction deduction, enabling the reverse heat conduction deduction to simultaneously incorporate the material's thermal conductivity properties, the heat transfer boundary location, and the relationship between the structural spatial coordinate distance.

[0083] Preferably, when performing reverse heat conduction deduction using the unsteady-state heat conduction deduction rule set, the surface temperature boundary in the thermal boundary condition of the target power equipment is first used as the starting point for deduction. The outer surface temperature profile corresponding to the surface temperature boundary in the three-dimensional spatial temperature distribution profile is read. Then, according to the determination rule of heat attenuation with spatial distance, the temperature attenuation relationship is traced back from the outer surface temperature profile to the interlayer heat transfer boundary corresponding to the adjacent heat transfer structure layer to form an interlayer backtracking temperature record. Subsequently, according to the heat diffusion rule at the junction of the structure layers, the interlayer backtracking temperature record is extended to the internal heat transfer boundary to form an internal backtracking temperature record. The internal backtracking temperature record is further combined with the heat transfer rule along the material's thermal conduction direction to organize the material's thermal conduction direction, so that the reverse heat conduction deduction can trace back layer by layer from the surface temperature boundary to the internal heat transfer structure layer.

[0084] Preferably, the internal heat conduction path is formed based on the reverse heat conduction deduction. Specifically, the interlayer backtracking temperature record and the internal backtracking temperature record are first mapped to the structural space coordinates in the three-dimensional structure of the target power equipment, and then the structural space coordinates with a continuous temperature decay relationship are connected according to the material's thermal conduction direction to form candidate internal heat transfer paths. Subsequently, based on the heat diffusion rules at the structural layer boundaries, it is determined whether the candidate internal heat transfer paths have a continuous diffusion relationship between adjacent heat transfer structural layers and internal heat conduction structural layers, and candidate internal heat transfer paths with a continuous diffusion relationship are retained to reconstruct the internal heat conduction path. The internal heat conduction path is used to express the spatial channel in the three-dimensional structure of the target power equipment from the surface temperature boundary back to the internal heat conduction structural layer, and the internal heat conduction path continues to serve as the path basis for temperature field reconstruction processing.

[0085] Preferably, when performing temperature field reconstruction processing along the internal heat conduction path, the structural spatial coordinates, material thermal conductivity properties, interlayer backtracking temperature records, and internal backtracking temperature records on the internal heat conduction path are first read. Then, based on the structural spatial coordinates, the spatial orientation of the internal heat conduction path within the three-dimensional structure of the target power equipment is determined. Based on the material thermal conductivity properties, the material thermal conductivity difference between adjacent structural spatial coordinates on the internal heat conduction path is determined. Based on the interlayer backtracking temperature records and the internal backtracking temperature records, the temperature variation relationship of the internal heat conduction path is determined. Subsequently, the spatial orientation of the internal heat conduction path, the material thermal conductivity difference of the internal heat conduction path, and the temperature variation relationship of the internal heat conduction path are correspondingly organized to form a path temperature reconstruction record. This path temperature reconstruction record continues to participate in the temperature field reconstruction processing, enabling the temperature field reconstruction processing to unfold along the internal heat conduction path within the three-dimensional structure of the target power equipment.

[0086] Preferably, the temperature field reconstruction process does not merely perform spatial interpolation on the surface temperature, but rather fills in the path temperature reconstruction record layer by layer with the structural spatial coordinates in the three-dimensional structure of the target power equipment. Specifically, first, the temperature variation relationship of the internal heat conduction path in the path temperature reconstruction record is written into the structural spatial coordinates where the internal heat conduction path is located. Then, according to the heat diffusion rules at the structural layer boundaries, the adjacent structural spatial coordinates around the internal heat conduction path are included in the temperature field reconstruction processing range. Subsequently, according to the determination rules of heat attenuation with spatial distance, the adjacent structural spatial coordinates within the temperature field reconstruction processing range are assigned corresponding reconstruction temperature values ​​to form a three-dimensional temperature field reconstruction record. The three-dimensional temperature field reconstruction record expresses the reconstruction temperature values ​​corresponding to multiple structural spatial coordinates in the three-dimensional structure of the target power equipment, and the three-dimensional temperature field reconstruction record continues to participate in the extraction of three-dimensional isothermal surface geometric features.

[0087] Preferably, when extracting the geometric features of the three-dimensional isothermal surface, the structural spatial coordinates corresponding to the reconstructed temperature values ​​belonging to the same temperature level are first found from the three-dimensional temperature field reconstruction record. Then, the structural spatial coordinates corresponding to the reconstructed temperature values ​​belonging to the same temperature level are connected according to the spatial adjacency relationship in the three-dimensional structure of the target power equipment to form candidate isothermal spatial surfaces. Subsequently, the candidate isothermal spatial surfaces are structurally hierarchically organized according to the internal heat conduction path and the thermal conductivity properties of the material to extract the geometric features of the three-dimensional isothermal surface. The geometric features of the three-dimensional isothermal surface include the spatial location of the isothermal surface, the extension direction of the isothermal surface, the covering structural layer of the isothermal surface, and the temperature level of the isothermal surface. The spatial location of the isothermal surface is used to express the spatial attachment area of ​​the candidate isothermal spatial surface in the three-dimensional structure of the target power equipment. The extension direction of the isothermal surface is used to express the directional relationship of the candidate isothermal spatial surface along the internal heat conduction path. The covering structural layer of the isothermal surface is used to express the target power equipment outer surface, adjacent heat transfer structural layer, and internal heat conduction structural layer through which the candidate isothermal spatial surface passes. The temperature level of the isothermal surface is used to express the range of reconstructed temperature values ​​corresponding to the candidate isothermal spatial surface.

[0088] Preferably, when encapsulating a three-dimensional spatial structure based on the geometric features of the three-dimensional isothermal surface, the spatial position, extension direction, covering structure layer, and temperature level of the isothermal surface are first read from the geometric features. Then, the spatial position of the isothermal surface is bound to the structural spatial coordinates of the target power equipment's three-dimensional structure; the covering structure layer is bound to the thermal conductivity properties of the material; and the temperature level of the isothermal surface is bound to the reconstructed temperature value in the three-dimensional temperature field reconstruction record, thus forming a three-dimensional temperature field encapsulation record. This three-dimensional temperature field encapsulation record is further spatially organized according to the structural spatial coordinates of the target power equipment's three-dimensional structure, enabling the record to simultaneously express temperature distribution, spatial position, structural hierarchy, and material thermal conductivity differences.

[0089] Preferably, the three-dimensional temperature field distribution map of the target power equipment is jointly constructed by the three-dimensional temperature field reconstruction record, the three-dimensional isothermal surface geometric features, the three-dimensional temperature field encapsulation record, and the internal heat conduction path. Specifically, the three-dimensional temperature field reconstruction record provides the reconstructed temperature values ​​corresponding to the spatial coordinates of each structure within the three-dimensional structure of the target power equipment; the three-dimensional isothermal surface geometric features provide the spatial surface morphology corresponding to the candidate isothermal spatial surfaces; the three-dimensional temperature field encapsulation record provides the binding relationship between the reconstructed temperature values ​​and the structural spatial coordinates and the thermal conductivity properties of the material; and the internal heat conduction path provides the spatial channel tracing back from the surface temperature boundary to the internal thermally conductive structural layer, as well as the heat transfer correspondence between the internal thermally conductive structural layer and the outer surface of the target power equipment. The three-dimensional temperature field reconstruction record, the three-dimensional isothermal surface geometric features, the three-dimensional temperature field encapsulation record, and the internal heat conduction path together form the three-dimensional temperature field distribution map of the target power equipment, enabling the map to express the temperature distribution relationship between the interior and exterior surfaces of the target power equipment, and to continue serving as the basis for subsequent location of candidate heat-generating areas.

[0090] Preferably, in the power distribution equipment inspection scenario, when the three-dimensional structure of the target power equipment includes the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end, firstly, based on the thermal boundary conditions of the target power equipment corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end, respectively, the unsteady-state heat conduction deduction rule sets corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end are respectively invoked; then, reverse heat conduction deduction is performed through the unsteady-state heat conduction deduction rule sets corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end to reconstruct the internal heat conduction paths corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end; subsequently, temperature field reconstruction processing is performed along the internal heat conduction paths corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end to extract the three-dimensional isothermal surface geometric features corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end. Based on the three-dimensional isothermal surface geometry corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal, a three-dimensional temperature field distribution map of the target power equipment is constructed that can distinguish the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal.

[0091] Optionally, the step of performing fault confirmation processing on the candidate heating area and generating fault confirmation results for the candidate heating area includes: The preset range is determined based on the spatial adjacency relationship between the candidate heat-generating area and adjacent structures in the three-dimensional structure of the target power equipment; Track the heat decay curve of the candidate heat-generating region radiating to adjacent structures within the preset range; A trend fitting analysis is performed on the heat decay curve to determine the heat generation trend characteristics; Obtain a pre-constructed feature set of non-faulty environment heat sources based on non-faulty environment heat source samples; The heat generation trend features are compared with the heat source feature set of the non-faulty environment to prevent false alarms, and the fault confirmation results of the candidate heat generation area are generated based on the false alarm prevention cross-comparison results.

[0092] Preferably, before performing fault confirmation processing on the candidate heating area, the structural spatial coordinates, reconstructed temperature value, three-dimensional isothermal surface geometric features, and internal heat conduction path corresponding to the candidate heating area are first read from the three-dimensional temperature field distribution map of the target power equipment. The structural spatial coordinates, reconstructed temperature value, three-dimensional isothermal surface geometric features, and internal heat conduction path corresponding to the candidate heating area are then organized to form a spatial positioning record of the candidate heating area. The spatial positioning record of the candidate heating area is used to express the spatial position, temperature distribution state, isothermal surface coverage state, and heat transfer source direction of the candidate heating area in the three-dimensional structure of the target power equipment. The spatial position is determined by the structural spatial coordinates corresponding to the candidate heating area, the temperature distribution state is determined by the reconstructed temperature value, the isothermal surface coverage state is determined by the three-dimensional isothermal surface geometric features, and the heat transfer source direction is determined by the internal heat conduction path. The spatial positioning record of the candidate heating area continues to participate in the determination of the spatial adjacency relationship between the candidate heating area and the adjacent structure, so that the subsequent fault confirmation process no longer judges based solely on the reconstructed temperature value of the candidate heating area itself, but combines the spatial position, temperature distribution, isothermal surface coverage, and heat transfer source direction of the candidate heating area in the three-dimensional structure of the target power equipment.

[0093] Preferably, when determining the preset range based on the spatial adjacency relationship between the candidate heat-generating area and adjacent structures in the three-dimensional structure of the target power equipment, the outer surface of the target power equipment, adjacent heat transfer structure layers, and internal heat conduction structure layers surrounding the candidate heat-generating area are first read according to the spatial positioning record of the candidate heat-generating area, and the outer surface of the target power equipment, adjacent heat transfer structure layers, and internal heat conduction structure layers surrounding the candidate heat-generating area are determined as adjacent structures to be judged; then, based on the structural spatial coordinates in the three-dimensional structure of the target power equipment, it is determined whether there is a spatial adjacency relationship between the adjacent structures to be judged and the candidate heat-generating area, so as to filter out the adjacent structures that have a spatial adjacency relationship with the candidate heat-generating area from the adjacent structures to be judged, and form a candidate heat-generating area adjacency structure record; wherein, the candidate heat-generating area adjacency structure record is used to express the adjacent structures around the candidate heat-generating area that can transfer heat, the structural spatial coordinates corresponding to the adjacent structures, and the structural hierarchy relationship between the adjacent structures and the candidate heat-generating area. The adjacent structure record of the candidate heating region is continuously read in correspondence with the thermal conductivity properties of the material to determine the type and direction of the thermally conductive material that the candidate heating region passes through when transferring heat to the adjacent structure, thereby providing structural and thermal basis for determining the preset range.

[0094] Preferably, the preset range is not a fixed two-dimensional image window, but a three-dimensional structure tracking range determined based on the adjacent structure record of the candidate heating region and the thermal conductivity properties of the material. Specifically, firstly, adjacent structures with a direct spatial adjacency to the candidate heating region are selected based on the adjacent structure record of the candidate heating region. Then, heat transfer adjacent structures with a heat transfer relationship to the candidate heating region are filtered from the adjacent structures based on the thermal conductivity properties of the material. Subsequently, the spatial coordinates of the structures corresponding to the heat transfer adjacent structures are organized into a preset range. The heat transfer adjacent structures are used to represent adjacent structures within the preset range that can receive the heat generated by the candidate heating region. The preset range is used to define the tracking boundary when subsequently tracking the heat decay curve emanating from the candidate heating region to the heat transfer adjacent structures. The preset range continues to serve as a spatial constraint for sampling the heat decay curve, enabling the heat decay curve to expand around the candidate heating region within the heat transfer neighborhood defined by the preset range in the three-dimensional structure of the target power equipment.

[0095] Preferably, when tracking the heat decay curve of the candidate heating region radiating to adjacent heat transfer structures within the preset range, the reconstructed temperature values ​​corresponding to the spatial coordinates of each structure in the candidate heating region are first compared, and the spatial coordinate of the structure with the largest reconstructed temperature value in the candidate heating region is determined as the starting point for heat decay tracking. Then, the spatial adjacency relationship expressed by the adjacent structures of the candidate heating region is recorded, and the reconstructed temperature value, structural spatial coordinates, and material thermal conductivity properties corresponding to each adjacent heat transfer structure within the preset range are read sequentially to form heat decay sampling nodes. The heat decay sampling nodes are used to express the temperature response, spatial distance relationship, and thermal conductivity material relationship at each adjacent heat transfer structure location when the candidate heating region radiates to the adjacent heat transfer structure. The heat decay sampling nodes are further arranged in order extending from the starting point of heat decay tracking to the outer edge of the preset range to form a heat decay curve, so that the heat decay curve can express the decay state formed by the temperature changing with spatial adjacency relationship and material thermal conductivity properties during the radiation of the candidate heating region to the adjacent heat transfer structure.

[0096] Preferably, the heat decay curve includes a heat decay tracking start point, multiple heat decay sampling nodes, a reconstructed temperature value corresponding to each heat decay sampling node, a structural spatial coordinate distance order corresponding to each heat decay sampling node, and a material thermal conductivity property corresponding to each heat decay sampling node; wherein, the structural spatial coordinate distance order is used to express the spatial extension order of the heat decay sampling node relative to the heat decay tracking start point, and the material thermal conductivity property is used to express the heat transfer capacity of the structural layer to which the heat decay sampling node belongs. The heat decay curve continues to participate in trend fitting analysis, so that the trend fitting analysis can simultaneously read the heat decay tracking start point, the heat decay sampling nodes, the reconstructed temperature value, the structural spatial coordinate distance order, and the material thermal conductivity property, and perform fitting based on the decreasing relationship of the reconstructed temperature value between the heat decay sampling nodes, the extension relationship of the structural spatial coordinate distance order, and the difference relationship of the material thermal conductivity property, rather than performing isolated fitting only on temperature values ​​that lack spatial correlation.

[0097] Preferably, when performing trend fitting analysis on the heat decay curve, the heat decay curve is first divided into a nearest neighbor curve segment close to the candidate heating region, an intermediate curve segment located in the middle of the adjacent heat transfer structure, and an outer edge curve segment close to the outer edge of the preset range, according to the spatial coordinate distance of the structure. Then, the reconstructed temperature value change relationship in the nearest neighbor curve segment, the intermediate curve segment, and the outer edge curve segment are read respectively, and the reconstructed temperature value change relationship is correspondingly organized with the thermal conductivity properties of the material to form a segmented heat decay organization record. The segmented heat decay organization record is used to express the temperature decay state of the nearest neighbor curve segment, the intermediate curve segment, and the outer edge curve segment under different material thermal conductivity properties. The segmented heat decay organization record is further used to determine whether the heat decay curve shows a heat transfer trend of gradual decay from the candidate heating region to the adjacent heat transfer structure, so that the trend fitting analysis can reflect the spatial propagation state of heat in the three-dimensional structure of the target power equipment.

[0098] Preferably, the heating trend characteristics are formed based on trend fitting analysis. Specifically, the temperature decay direction, temperature decay continuity, temperature decay amplitude variation relationship, and material thermal conductivity property correlation relationship are first extracted from the segmented heat decay records. Then, the temperature decay direction, temperature decay continuity, temperature decay amplitude variation relationship, and material thermal conductivity property correlation relationship are combined and organized to determine the heating trend characteristics. The temperature decay direction is used to express the direction of temperature decrease from the heat decay tracking starting point to the outer edge of the preset range. The temperature decay continuity is used to express whether there is a continuous temperature change relationship between adjacent heat decay sampling nodes. The temperature decay amplitude variation relationship is used to express the difference in temperature decrease amplitude between the nearest curve segment, the middle curve segment, and the outer curve segment. The material thermal conductivity property correlation relationship is used to express the correspondence between the temperature decay direction, temperature decay continuity, temperature decay amplitude variation relationship, and material thermal conductivity property. The heat generation trend feature is used to express whether the candidate heat generation area has a tendency to transfer heat from the internal heat conduction structure layer or the adjacent heat transfer structure layer to the outer surface of the target power equipment. The heat generation trend feature continues to participate in the subsequent false alarm prevention cross-comparison, so that the fault confirmation result of the candidate heat generation area can be generated based on the difference between the spatial heat transfer trend and the heat source characteristics of the non-fault environment.

[0099] Preferably, before acquiring the non-faulty environmental heat source feature set pre-constructed based on the non-faulty environmental heat source samples, a non-faulty environmental heat source sample is first constructed. This sample originates from environmental heat source manifestations during the inspection of the target power equipment that do not involve internal equipment faults causing heat generation. These environmental heat source manifestations include surface temperature rises caused by external environmental radiation, localized temperature rises caused by reflections from surrounding structures, and gradual surface temperature changes caused by short-term changes in heat dissipation conditions. When organizing the non-faulty environmental heat source samples, the non-faulty spatial location, non-faulty temperature change relationship, non-faulty heat attenuation pattern, and non-faulty perspective change response corresponding to each sample are first read. Then, the non-faulty spatial location, non-faulty temperature change relationship, non-faulty heat attenuation pattern, and non-faulty perspective change response are saved accordingly to form a non-faulty environmental heat source sample record. The non-faulty environment heat source sample record is used to save the spatial source, temperature change source, heat decay source, and viewpoint response source of the non-faulty environment heat source sample. The non-faulty environment heat source sample record continues to participate in the construction of the non-faulty environment heat source feature set, so that the non-faulty environment heat source feature set can serve as a reference basis for cross-comparison to prevent false alarms.

[0100] Preferably, the non-faulty environment heat source feature set is pre-constructed based on the non-faulty environment heat source sample records. Specifically, non-faulty temperature gradual change features, non-faulty spatial adhesion features, non-faulty perspective synchronous change features, and non-faulty material thermal conductivity weak correlation features are first extracted from the non-faulty environment heat source sample records. Then, the non-faulty temperature gradual change features, the non-faulty spatial adhesion features, the non-faulty perspective synchronous change features, and the non-faulty material thermal conductivity weak correlation features are structured and organized to form the non-faulty environment heat source feature set. The non-faulty temperature gradual change features are used to express the slow change relationship of the temperature corresponding to the non-faulty environment heat source sample with the change of inspection time or perspective. The non-faulty spatial adhesion features are used to express the spatial state in which the temperature corresponding to the non-faulty environment heat source sample is not stably attached to the internal thermally conductive structural layer or adjacent heat transfer structural layer. The non-faulty perspective synchronous change features are used to express the state in which the temperature corresponding to the non-faulty environment heat source sample changes synchronously with the change of observation perspective. The non-faulty material thermal conductivity weak correlation features are used to express the state in which the temperature corresponding to the non-faulty environment heat source sample has a low degree of correlation with the thermal conductivity properties of the material. The non-faulty environment heat source feature set is then cross-compared with the heating trend feature to prevent false alarms.

[0101] Preferably, when performing a false alarm-prevention cross-comparison between the heating trend feature and the non-faulty environment heat source feature set, the temperature decay direction in the heating trend feature is first compared with the non-faulty spatial attachment feature in the non-faulty environment heat source feature set to determine whether the candidate heating region has a spatial attachment relationship with the internal heat-conducting structure layer or adjacent heat transfer structure layer; then, the temperature decay continuity in the heating trend feature is compared with the non-faulty temperature gradual change feature in the non-faulty environment heat source feature set to determine whether the candidate heating region exhibits a heat transfer state of continuous decay along the preset range; subsequently, the material thermal conductivity property correlation in the heating trend feature is compared with the non-faulty material thermal conductivity weak correlation feature in the non-faulty environment heat source feature set to determine whether the heat decay curve of the candidate heating region is affected by the material thermal conductivity property; and the viewing angle change state corresponding to the heating trend feature is compared with the non-faulty viewing angle synchronous change feature in the non-faulty environment heat source feature set to determine whether the candidate heating region mainly changes with the viewing angle. The comparison of spatial attachment relationships, the comparison of heat transfer states, the comparison of the influence of material thermal conductivity properties, and the comparison of changes in observation perspective together form a false alarm prevention cross-comparison result. The false alarm prevention cross-comparison result continues to participate in the generation of the candidate heating area fault confirmation result.

[0102] Preferably, when generating the fault confirmation result of the candidate heating area based on the false alarm cross-comparison result, the comparison content regarding spatial attachment relationship, continuous attenuation relationship, material thermal conductivity property correlation relationship, and observation perspective change relationship in the false alarm cross-comparison result is first read; when the false alarm cross-comparison result indicates that the candidate heating area has a spatial attachment relationship with the internal thermal conductive structure layer or adjacent heat transfer structure layer, the heat attenuation curve has a continuous attenuation relationship within the preset range, and the heating trend feature has a correlation with the material thermal conductivity property, the candidate heating area is marked as a candidate heating area that has passed the fault confirmation process, and a fault confirmation result representing the candidate heating area that has passed the fault confirmation process is generated; when the false alarm cross-comparison result indicates that the candidate heating area matches the non-fault perspective synchronous change feature or the non-fault material thermal conductivity weak correlation feature in the non-fault environment heat source feature set, the candidate heating area is marked as a candidate heating area that has not passed the fault confirmation process, and a fault confirmation result representing the candidate heating area that has not passed the fault confirmation process is generated. The fault confirmation results of the candidate heating areas continue to participate in the subsequent absolute heat quantification, so that the subsequent absolute heat quantification can be carried out for the candidate heating areas that have passed the fault confirmation process, and the candidate heating areas that have not passed the fault confirmation process will not enter the subsequent absolute heat quantification.

[0103] Preferably, in the scenario of inspecting power distribution equipment, when the candidate heat-generating area is located on the outer surface of the enclosure, the outer surface of the heat sink, or the outer surface of the wiring terminal, the adjacent structures corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal are first determined according to the three-dimensional structure of the target power equipment. Then, based on the adjacent structures corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal, preset ranges corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal are determined respectively. Finally, the candidate heat-generating area corresponding to the outer surface of the enclosure is traced towards the corresponding area on the outer surface of the enclosure. The system generates heat decay curves emanating from adjacent structures within a preset range, heat decay curves emanating from candidate heat-generating areas on the outer surface of the heat sink towards adjacent structures within a preset range, and heat decay curves emanating from candidate heat-generating areas on the outer surface of the wiring terminals towards adjacent structures within a preset range. Subsequently, trend fitting analysis is performed on the heat decay curves corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals to determine the heat trend characteristics corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals. These heat trend characteristics are then cross-compared with the non-fault environment heat source feature set to prevent false alarms, generating fault confirmation results for candidate heat-generating areas that can distinguish between the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminals.

[0104] Optionally, the step of quantifying the absolute heat of the candidate heating region to determine the target heat dissipation and the absolute temperature rise parameter of the candidate heating region includes: Determine the current-carrying cross-sectional area parameters and estimated operating current values ​​of the candidate heating region in the three-dimensional structure of the target power equipment; Based on the current-carrying cross-sectional area parameter and the estimated operating current value, an internal resistance heating quantification rule is established. The internal resistance heating quantification rule includes a judgment rule that the smaller the current-carrying cross-sectional area, the higher the resistance heating contribution; a judgment rule that the larger the estimated operating current value, the higher the resistance heating contribution; and a quantification rule corresponding to the heat accumulation state and heat dissipation. Discrete sampling is performed on the temperature field corresponding to the candidate heating region in the three-dimensional temperature field distribution map of the target power equipment to obtain discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region. The discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region are substituted into the internal resistance heating quantification rule for thermoelectric coupling iterative matching processing to determine the target heating power consumption. The absolute temperature rise parameter of the candidate heating region is determined based on the discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region and the thermal boundary conditions of the target power equipment.

[0105] Preferably, before performing absolute heat quantification on the candidate heating regions, the fault confirmation results of the candidate heating regions, the spatial positioning records of the candidate heating regions, the three-dimensional temperature field distribution map of the target power equipment, and the thermal boundary conditions of the target power equipment are read first. The fault confirmation results of the candidate heating regions are used to define the candidate heating regions that can be included in the absolute heat quantification. The spatial positioning records of the candidate heating regions are used to express the spatial location, temperature distribution, isothermal surface coverage, and heat transfer source direction of the candidate heating regions within the three-dimensional structure of the target power equipment. The three-dimensional temperature field distribution map of the target power equipment is used to provide reconstructed temperature values ​​for the candidate heating regions and their adjacent spatial locations. The thermal boundary conditions of the target power equipment are used to provide surface temperature boundaries, interlayer heat transfer boundaries, and internal heat transfer boundaries. After reading the fault confirmation results of the candidate heating areas, the candidate heating areas that have passed the fault confirmation process are taken as the absolute heat quantification areas. The absolute heat quantification areas are then read in correspondence with the spatial positioning records of the candidate heating areas, the three-dimensional temperature field distribution map of the target power equipment, and the thermal boundary conditions of the target power equipment to form the absolute heat quantification basic reading records. The absolute heat quantification basic reading records continue to participate in the determination of current-carrying cross-sectional area parameters, estimated operating current values, three-dimensional temperature field discrete values, target heat dissipation power consumption, and absolute temperature rise parameters of the candidate heating areas, so that the above processing results all revolve around the same absolute heat quantification area.

[0106] Preferably, when determining the current-carrying cross-sectional area parameter of the candidate heating region in the three-dimensional structure of the target power equipment, the structural spatial coordinates and internal heat conduction path corresponding to the absolute heat quantification region are first read according to the spatial positioning record of the candidate heating region in the absolute heat quantification basic reading record. Then, the current-carrying channel corresponding to the absolute heat quantification region is found in the three-dimensional structure of the target power equipment according to the structural spatial coordinates. The current-carrying channel is used to express the position of the conductive structure in the target power equipment that carries the current corresponding to the estimated operating current and has a thermal relationship with the absolute heat quantification region. Subsequently, the cross-sectional boundary position of the current-carrying channel adjacent to the absolute heat quantification region is determined according to the internal heat conduction path, and the cross-sectional geometry is adjusted according to the structural spatial coordinates of the cross-sectional boundary position in the three-dimensional structure of the target power equipment to form a current-carrying structure cross-sectional record. The current-carrying structure cross-sectional record includes the cross-sectional boundary position of the current-carrying channel, the cross-sectional boundary shape, and the thermal conductivity properties of the material to which the cross-sectional boundary position belongs. The current-carrying cross-sectional area parameter is determined based on the current-carrying structure cross-section record, so that the current-carrying cross-sectional area parameter can express the effective current-carrying cross-section of the current-carrying channel where the absolute heat quantification region is located, rather than just expressing the area of ​​the heat-generating region in the two-dimensional image.

[0107] Preferably, when determining the estimated operating current value of the candidate heat-generating region in the three-dimensional structure of the target power equipment, the current-carrying channel corresponding to the absolute heat quantification region, the operating condition of the target power equipment, and the thermal boundary condition of the target power equipment corresponding to the absolute heat quantification region are first read. The operating condition of the target power equipment expresses the load state of the target power equipment during inspection, and the thermal boundary condition of the target power equipment corresponding to the absolute heat quantification region expresses the temperature transfer boundary in the structural space where the absolute heat quantification region is located. Based on the correspondence between the current-carrying channel and the operating condition of the target power equipment, the operating current change record of the current-carrying channel under the operating condition of the target power equipment is extracted, and the operating current change record is correspondingly organized with the thermal boundary condition of the target power equipment corresponding to the absolute heat quantification region to form an operating current estimation record. The operating current estimation record expresses the current carrying state of the current-carrying channel where the absolute heat quantification region is located under the operating condition of the target power equipment. The estimated operating current value is determined based on the operating current estimation record, so that the estimated operating current value corresponds to the current-carrying channel where the absolute heat quantification region is located, the operating condition of the target power equipment, and the thermal boundary condition of the target power equipment.

[0108] Preferably, when establishing the internal resistance heating quantification rule, the current-carrying cross-sectional area parameter, the estimated operating current value, the material thermal conductivity property, the target power equipment thermal boundary condition corresponding to the absolute heat quantification region, and the spatial location record of the candidate heating region are first organized to form a basic record for internal resistance heating quantification. This basic record expresses the effective current-carrying cross-section, current carrying state, material thermal conductivity difference, heat transfer boundary, and spatial attachment position of the current-carrying channel where the absolute heat quantification region is located. Subsequently, the internal resistance heating quantification rule is established based on this basic record, enabling the rule to simultaneously read the current-carrying cross-sectional area parameter, the estimated operating current value, the material thermal conductivity property, and the target power equipment thermal boundary condition, rather than comparing the temperature value in isolation with a preset upper limit.

[0109] Preferably, the internal resistance heating quantification rules include a rule that a smaller current-carrying cross-sectional area results in a higher resistance heating contribution, a rule that a larger estimated operating current results in a higher resistance heating contribution, and a quantification rule corresponding to the heat accumulation state and heating power consumption. Specifically, the rule that a smaller current-carrying cross-sectional area results in a higher resistance heating contribution expresses that, under the same estimated operating current and the same material thermal conductivity, a smaller current-carrying cross-sectional area means a more concentrated current flow per unit cross-section in the current-carrying channel, resulting in a higher resistance heating contribution. The rule that a larger estimated operating current means a higher resistance heating contribution expresses that, under the same current-carrying cross-sectional area and the same material thermal conductivity, a larger estimated operating current means a stronger heat generation trend caused by the current in the current-carrying channel, resulting in a higher resistance heating contribution. The quantification rule corresponding to the heat accumulation state and heating power consumption expresses that, under the heat transfer boundary defined by the thermal boundary conditions of the target power equipment, when the reconstructed temperature value in the absolute heat quantification region and its adjacent structures continuously rises and the temperature decay range is wide, the corresponding heating power consumption is high. The determination rule that the smaller the current-carrying cross-sectional area, the higher the contribution of resistive heating; the determination rule that the larger the estimated operating current value, the higher the contribution of resistive heating; and the quantification rule corresponding to the heat accumulation state and heat dissipation power consumption together constitute the internal resistance heating quantification rule. The internal resistance heating quantification rule continues to participate in the thermoelectric coupling iterative matching processing of the discrete values ​​of the three-dimensional temperature field.

[0110] Preferably, when discretely sampling the temperature field corresponding to the candidate heating region in the three-dimensional temperature field distribution map of the target power equipment, the spatial boundary of the absolute heat quantification region in the three-dimensional temperature field distribution map of the target power equipment is first determined according to the spatial positioning record of the candidate heating region in the absolute heat quantification basic reading record. Then, according to the internal heat conduction path and the geometric features of the three-dimensional isothermal surface, the spatial coordinates of the structure within the spatial boundary and adjacent extensions along the internal heat conduction path are determined as temperature sampling positions. Subsequently, the reconstructed temperature value, material thermal conductivity properties, internal heat conduction path position, and boundary distance relationship with the thermal boundary condition of the target power equipment corresponding to each temperature sampling position are read from the three-dimensional temperature field distribution map of the target power equipment to form a temperature sampling record. The temperature sampling record is further arranged according to the spatial adjacency relationship of the temperature sampling position in the three-dimensional structure of the target power equipment to obtain the discrete three-dimensional temperature field value corresponding to the candidate heating region.

[0111] Preferably, the discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region include multiple temperature sampling locations, a reconstructed temperature value corresponding to each temperature sampling location, a material thermal conductivity property corresponding to each temperature sampling location, an internal heat conduction path location corresponding to each temperature sampling location, and a boundary distance relationship corresponding to each temperature sampling location. Specifically, the temperature sampling location represents the spatial location of the discrete sampling within the three-dimensional structure of the target power equipment; the reconstructed temperature value represents the temperature state at the temperature sampling location; the material thermal conductivity property represents the thermal conductivity difference between the structural layers to which the temperature sampling location belongs; the internal heat conduction path location represents the spatial relationship between the temperature sampling location and the internal heat conduction path; and the boundary distance relationship represents the distance between the temperature sampling location and the surface temperature boundary, interlayer heat transfer boundary, and internal heat transfer boundary. The discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region are further substituted into the internal resistance heating quantification rule, enabling subsequent thermoelectric coupling iterative matching processing to perform joint calculations based on spatial location, temperature state, material thermal conductivity property, and heat transfer boundary.

[0112] Preferably, when substituting the discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region into the internal resistance heating quantification rule for thermoelectric coupling iterative matching processing, the current-carrying cross-sectional area parameter and the estimated operating current value are first used as resistance heating constraints, and the thermal conductivity properties of the material and the thermal boundary conditions of the target power equipment are used as heat transfer constraints to establish a candidate heating power consumption estimation record. The candidate heating power consumption estimation record is used to express the temperature distribution estimation state of the absolute heat quantification region in the three-dimensional structure of the target power equipment under different heating power consumption conditions. Subsequently, the temperature distribution estimation state corresponding to the candidate heating power consumption estimation record is matched with the discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region in round by round. After each round of matching, the candidate heating power consumption estimation record is adjusted according to the judgment rule that the smaller the current-carrying cross-sectional area, the higher the resistance heating contribution; the judgment rule that the larger the estimated operating current value, the higher the resistance heating contribution; and the quantification rule corresponding to the heat accumulation state and heating power consumption, so as to form a thermoelectric coupling iterative matching record. The thermoelectric coupling iterative matching record is then used to determine the target heating power consumption.

[0113] Preferably, the thermoelectric coupling iterative matching process does not directly multiply the estimated operating current value by the temperature value, nor does it simply compare the current-carrying cross-sectional area parameter with the reconstructed temperature value. Instead, it compares the reconstructed temperature value at each temperature sampling position in the discrete three-dimensional temperature field corresponding to the candidate heating region with the temperature distribution estimation state at the same temperature sampling position in the candidate heating power consumption estimation record. When the temperature distribution estimation state is lower than the reconstructed temperature value distribution expressed by the discrete three-dimensional temperature field corresponding to the candidate heating region, the heating power consumption estimation amount in the candidate heating power consumption estimation record is increased according to the internal resistance heating quantization rule. When the temperature distribution estimation state is higher than the reconstructed temperature value distribution expressed by the discrete three-dimensional temperature field corresponding to the candidate heating region, the heating power consumption estimation amount in the candidate heating power consumption estimation record is decreased according to the internal resistance heating quantization rule. When the difference between the temperature distribution estimation state and the discrete three-dimensional temperature field corresponding to the candidate heating region is within a preset matching range, the heating power consumption estimation amount in the candidate heating power consumption estimation record is determined as the target heating power consumption. The preset matching range is pre-configured based on the range of reconstructed temperature values ​​in the temperature sampling records and the thermal boundary conditions of the target power equipment. The target heat dissipation continues to participate in the determination of the absolute temperature rise parameter of the candidate heat dissipation area, so that the absolute temperature rise parameter of the candidate heat dissipation area can form a corresponding relationship with the target heat dissipation.

[0114] Preferably, when determining the absolute temperature rise parameter of the candidate heating region based on the discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region and the thermal boundary conditions of the target power equipment, the reconstructed temperature value distribution within the coverage area of ​​the absolute heat quantification region is first read from the discrete values ​​of the three-dimensional temperature field corresponding to the candidate heating region. Then, the surface temperature boundary, interlayer heat transfer boundary, and internal heat transfer boundary corresponding to the absolute heat quantification region are read from the thermal boundary conditions of the target power equipment. Subsequently, the boundary temperature reference on the outer surface side of the absolute heat quantification region is determined based on the surface temperature boundary, the boundary temperature reference at the junction of the structural layers of the absolute heat quantification region is determined based on the interlayer heat transfer boundary, and the boundary temperature reference on the inner side of the absolute heat quantification region is determined based on the internal heat transfer boundary. The reconstructed temperature value distribution within the coverage area of ​​the absolute heat quantification region is then compared with the boundary temperature references corresponding to the surface temperature boundary, the interlayer heat transfer boundary, and the internal heat transfer boundary to form a candidate heating region temperature rise adjustment record. This candidate heating region temperature rise adjustment record is then used to determine the absolute temperature rise parameter of the candidate heating region.

[0115] Preferably, the absolute temperature rise parameter of the candidate heating region is determined based on the temperature rise compilation record of the candidate heating region. Specifically, the temperature rise amplitude of each temperature sampling position within the coverage area of ​​the absolute heat quantification region relative to the surface temperature boundary, the interlayer heat transfer boundary, and the internal heat transfer boundary is first read from the temperature rise compilation record of the candidate heating region. Then, based on the location of the internal heat conduction path and the thermal conductivity of the material, the temperature rise amplitude is spatially assigned to form an absolute temperature rise spatial distribution record. The absolute temperature rise spatial distribution record is used to express the temperature rise state of the absolute heat quantification region under different structural spatial coordinates, different material thermal conductivity properties, and different heat transfer boundaries. The absolute temperature rise parameter of the candidate heating region is determined based on the absolute temperature rise spatial distribution record, so that the absolute temperature rise parameter of the candidate heating region no longer only represents the difference between the surface maximum temperature and the empirical threshold, but also represents the three-dimensional temperature rise state of the candidate heating region in the three-dimensional structure of the target power equipment.

[0116] Preferably, after the target heat dissipation and the absolute temperature rise parameter of the candidate heat dissipation area are formed, the target heat dissipation, the absolute temperature rise parameter of the candidate heat dissipation area, the current carrying cross-sectional area parameter, the estimated operating current value, the discrete value of the three-dimensional temperature field corresponding to the candidate heat dissipation area, and the fault confirmation result of the candidate heat dissipation area are organized accordingly to form an absolute heat quantification record. The absolute heat quantification record is used to express the heat dissipation state, absolute temperature rise state, current carrying cross-sectional area state, current carrying capacity state, and three-dimensional temperature field discrete state of the candidate heat dissipation area after fault confirmation processing. The absolute heat quantification record continues to participate in the output of the hazard absolute temperature rise identification result, so that the hazard absolute temperature rise identification result can simultaneously reflect the target heat dissipation and absolute temperature rise parameter of the candidate heat dissipation area.

[0117] Preferably, in the scenario of power distribution equipment inspection, when the candidate heat-generating area is located on the outer surface of the enclosure, the outer surface of the heat sink, or the outer surface of the wiring terminal, the current-carrying channels corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal are first determined according to the three-dimensional structure of the target power equipment, and the current-carrying cross-sectional area parameters corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal are determined respectively. Then, the estimated operating current values ​​corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal are determined according to the operating conditions of the target power equipment. Subsequently, the local temperature fields corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal in the three-dimensional temperature field distribution diagram of the target power equipment are discretely sampled to obtain the discrete values ​​of the three-dimensional temperature field corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring terminal in the three-dimensional temperature field distribution diagram of the target power equipment, respectively. After substituting the discrete values ​​of the three-dimensional temperature field corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end into the corresponding internal resistance heating quantification rules and performing thermoelectric coupling iterative matching processing, the target heat dissipation power corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end are determined respectively. In combination with the corresponding target power equipment thermal boundary conditions, the absolute temperature rise parameters of the candidate heating areas corresponding to the outer surface of the enclosure, the outer surface of the heat sink, and the outer surface of the wiring end are determined respectively.

[0118] like Figure 3As shown in the figure, this is an embodiment of a power inspection infrared feature decoupling and hazard absolute temperature rise automatic identification device. The power inspection infrared feature decoupling and hazard absolute temperature rise automatic identification device includes a pose normalization processing module, an infrared feature decoupling module, a three-dimensional temperature field construction module, and a hazard absolute temperature rise identification result output module. The pose normalization processing module is configured to acquire the original multi-view infrared thermal image of the target power equipment and the pre-constructed three-dimensional structure of the target power equipment. The three-dimensional structure of the target power equipment includes the structural space coordinates and material thermal conductivity properties of the target power equipment. The pose normalization processing is performed on the original multi-view infrared thermal image to obtain the multi-view infrared thermal image data source. The infrared feature decoupling module is configured to obtain a pre-configured environmental radiation elimination matrix, use the environmental radiation elimination matrix to decouple the features of the multi-view infrared thermal image data source, generate an infrared feature map of the target power equipment, and generate a surface temperature mapping feature set of the target power equipment based on the infrared feature map of the target power equipment. The three-dimensional temperature field construction module is configured to structurally fuse the surface temperature mapping feature set of the target power equipment with the three-dimensional structure of the target power equipment, determine the thermal boundary conditions of the target power equipment, perform heat conduction deduction based on the thermal boundary conditions of the target power equipment, and construct a three-dimensional temperature field distribution map of the target power equipment. The hidden danger absolute temperature rise identification result output module is configured to locate candidate heating areas from the three-dimensional temperature field distribution map of the target power equipment, perform fault confirmation processing on the candidate heating areas, generate a fault confirmation result for the candidate heating areas, and after the candidate heating areas pass the fault confirmation processing, perform absolute heat quantification on the candidate heating areas to determine the target heat consumption and the absolute temperature rise parameter of the candidate heating areas, and output the hidden danger absolute temperature rise identification result based on the target heat consumption and the absolute temperature rise parameter of the candidate heating areas.

[0119] like Figure 4 As shown, an electronic device according to an embodiment of this application is provided. The electronic device includes: a memory, a processor, and a power inspection infrared feature decoupling and hidden danger absolute temperature rise automatic identification program stored in the memory and executable on the processor. The power inspection infrared feature decoupling and hidden danger absolute temperature rise automatic identification program is configured to implement the steps of the power inspection infrared feature decoupling and hidden danger absolute temperature rise automatic identification method as described in any one of the present application.

[0120] like Figure 5As shown, this is a computer-readable storage medium according to an embodiment of the present application. The computer-readable storage medium stores a program for automatic identification of absolute temperature rise of infrared features during power inspection and hidden dangers. When the program is executed by a processor, it implements the steps of the method for automatic identification of absolute temperature rise of infrared features during power inspection and hidden dangers as described in any one of the present application.

[0121] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A power patrol infrared feature decoupling and hidden danger absolute temperature rise automatic identification method, characterized in that, Includes the following steps: The original multi-view infrared thermal image of the target power equipment and the pre-constructed three-dimensional structure of the target power equipment are obtained. The three-dimensional structure of the target power equipment includes the structural spatial coordinates and material thermal conductivity properties of the target power equipment. The original multi-view infrared thermal image is subjected to pose normalization processing to obtain the multi-view infrared thermal image data source. A pre-configured environmental radiation rejection matrix is ​​obtained, and the multi-view infrared thermal image data source is decoupled using the environmental radiation rejection matrix to generate an infrared feature map of the target power equipment. A surface temperature mapping feature set of the target power equipment is then generated based on the infrared feature map of the target power equipment. The surface temperature mapping feature set of the target power equipment is structurally fused with the three-dimensional structure of the target power equipment to determine the thermal boundary conditions of the target power equipment. Based on the thermal boundary conditions of the target power equipment, heat conduction is deduced to construct a three-dimensional temperature field distribution map of the target power equipment. Candidate heating areas are located from the three-dimensional temperature field distribution map of the target power equipment. Fault confirmation processing is performed on the candidate heating areas to generate candidate heating area fault confirmation results. After the candidate heating area passes the fault confirmation processing, the absolute heat of the candidate heating area is quantified to determine the target heat consumption and the absolute temperature rise parameter of the candidate heating area. Based on the target heat consumption and the absolute temperature rise parameter of the candidate heating area, the hazard absolute temperature rise identification result is output.

2. The electric power patrol infrared feature decoupling and hidden danger absolute temperature rise automatic identification method according to claim 1, characterized in that, The step of performing pose normalization processing on the original multi-view infrared thermal image to obtain the multi-view infrared thermal image data source includes: Obtain the shooting pose parameters corresponding to the original multi-view infrared thermal image; Based on the shooting pose parameters, the original multi-view infrared thermal image is spatially projected and transformed, and the original multi-view infrared thermal image after spatial projection transformation is contour aligned according to the physical geometric contour in the three-dimensional structure of the target power equipment to obtain a pose-normalized thermal image sequence. The thermal intensity of the pose-normalized thermal image sequence is calibrated to obtain the multi-view infrared thermal image data source.

3. The power patrol infrared feature decoupling and hazard absolute temperature rise automatic identification method according to claim 2, characterized in that, The step of performing thermal intensity calibration on the pose-normalized thermal image sequence to obtain the multi-view infrared thermal image data source includes: Obtain the grayscale mapping features of pixels in the pose-normalized heatmap sequence; The grayscale mapping features are subjected to temperature inversion transformation according to the preset blackbody radiation reference calibration mapping relationship to obtain an absolute temperature distribution matrix sequence. The absolute temperature distribution matrix sequence is linked with the pose-normalized thermal image sequence to obtain the multi-view infrared thermal image data source.

4. The method for decoupling infrared features and automatically identifying absolute temperature rise of potential hazards during power line inspection as described in claim 1, characterized in that, The step of using the environmental radiation elimination matrix to decouple features from the multi-view infrared thermal image data source and generate an infrared feature map of the target power equipment includes: Obtain the pre-configured environmental radiation rejection matrix, wherein the pre-configured environmental radiation rejection matrix is ​​determined by the correspondence between the background radiation characteristics of the inspection environment and the infrared imaging channel, and the background radiation characteristics of the inspection environment and the correspondence between the infrared imaging channel jointly define the distribution of matrix elements in the pre-configured environmental radiation rejection matrix; The environmental radiation removal matrix is ​​applied to the multi-view infrared thermal image data source to establish an independent component analysis matrix. The energy weight distribution ratio of different heat source components is determined by performing blind source separation iterative calculation on the multi-view infrared thermal image data source using the independent component analysis matrix. The environmental interference component is determined based on the energy weight distribution ratio, and the environmental interference component is removed from the multi-view infrared thermal image data source to generate the infrared feature map of the target power equipment.

5. The method for decoupling infrared features and automatically identifying absolute temperature rise of potential hazards during power line inspection as described in claim 1, characterized in that, The step of generating a surface temperature mapping feature set of the target power equipment based on the infrared feature map of the target power equipment includes: Determine the thermal gradient features in the infrared feature map of the target power equipment; Obtain a preset temperature threshold that matches the device type and operating conditions of the target power equipment, wherein the preset temperature threshold is pre-configured based on the device type and operating conditions of the target power equipment; Spatial connectivity analysis is performed on the thermal gradient features to delineate target temperature connectivity regions with temperature values ​​greater than the preset temperature threshold. The contour boundary coordinates corresponding to the target temperature connected region are determined, and the contour boundary coordinates are structurally recombined with the infrared feature map of the target power equipment to generate the surface temperature mapping feature set of the target power equipment.

6. The method for decoupling infrared features and automatically identifying absolute temperature rise of potential hazards during power line inspection as described in claim 1, characterized in that, The step of structurally fusing the surface temperature mapping feature set of the target power equipment with the three-dimensional structure of the target power equipment to determine the thermal boundary conditions of the target power equipment includes: Obtain a preset physical depth level pre-configured according to the entity structure layer relationship in the three-dimensional structure of the target power equipment, wherein the entity structure layer relationship is used to define the layer position of the preset physical depth level in the three-dimensional structure of the target power equipment; The surface temperature mapping feature set of the target power equipment is hierarchically parsed according to the preset physical depth level in order to extract the surface heat dissipation gradient grid representing different preset physical depth levels. The surface heat dissipation gradient grid is mapped to the three-dimensional structure of the target power equipment in spatial coordinates to determine the three-dimensional spatial temperature distribution profile. The thermal boundary conditions of the target power equipment are determined based on the three-dimensional spatial temperature distribution profile.

7. The method for decoupling infrared features and automatically identifying absolute temperature rise of potential hazards during power line inspection as described in claim 1, characterized in that, The step of constructing a three-dimensional temperature field distribution map of the target power equipment by performing heat conduction deduction based on the thermal boundary conditions of the target power equipment includes: The set of unsteady-state heat conduction inference rules is called to match the material thermal conductivity properties in the three-dimensional structure of the target power equipment and the thermal boundary conditions of the target power equipment. The set of unsteady-state heat conduction inference rules includes rules for heat transfer along the material thermal conductivity direction, rules for heat diffusion at the junction of structural layers, and rules for determining the attenuation of heat with spatial distance. Using the unsteady-state heat conduction deduction rule set, reverse heat conduction deduction is performed based on the thermal boundary conditions of the target power equipment to reconstruct the internal heat conduction path; Temperature field reconstruction is performed along the internal heat conduction path to extract the geometric features of the three-dimensional isothermal surface; Based on the geometric features of the three-dimensional isothermal surface, a three-dimensional spatial structure is encapsulated to construct a three-dimensional temperature field distribution map of the target power equipment.

8. A device for decoupling infrared features and automatically identifying absolute temperature rise of potential hazards during power line inspections, characterized in that, The power patrol infrared feature decoupling and automatic hazard absolute temperature rise identification device includes a pose normalization processing module, an infrared feature decoupling module, a three-dimensional temperature field construction module, and a hazard absolute temperature rise identification result output module. The pose normalization processing module is configured to acquire the original multi-view infrared thermal image of the target power equipment and the pre-constructed three-dimensional structure of the target power equipment. The three-dimensional structure of the target power equipment includes the structural space coordinates and material thermal conductivity properties of the target power equipment. The pose normalization processing is performed on the original multi-view infrared thermal image to obtain the multi-view infrared thermal image data source. The infrared feature decoupling module is configured to obtain a pre-configured environmental radiation elimination matrix, use the environmental radiation elimination matrix to decouple the features of the multi-view infrared thermal image data source, generate an infrared feature map of the target power equipment, and generate a surface temperature mapping feature set of the target power equipment based on the infrared feature map of the target power equipment. The three-dimensional temperature field construction module is configured to structurally fuse the surface temperature mapping feature set of the target power equipment with the three-dimensional structure of the target power equipment, determine the thermal boundary conditions of the target power equipment, perform heat conduction deduction based on the thermal boundary conditions of the target power equipment, and construct a three-dimensional temperature field distribution map of the target power equipment. The hidden danger absolute temperature rise identification result output module is configured to locate candidate heating areas from the three-dimensional temperature field distribution map of the target power equipment, perform fault confirmation processing on the candidate heating areas, generate a fault confirmation result for the candidate heating areas, and after the candidate heating areas pass the fault confirmation processing, perform absolute heat quantification on the candidate heating areas to determine the target heat consumption and the absolute temperature rise parameter of the candidate heating areas, and output the hidden danger absolute temperature rise identification result based on the target heat consumption and the absolute temperature rise parameter of the candidate heating areas.

9. An electronic device, characterized in that, The electronic device includes: a memory, a processor, and a power inspection infrared feature decoupling and hidden danger absolute temperature rise automatic identification program stored in the memory and executable on the processor. The power inspection infrared feature decoupling and hidden danger absolute temperature rise automatic identification program is configured to implement the steps of the power inspection infrared feature decoupling and hidden danger absolute temperature rise automatic identification method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program for automatic identification of power inspection infrared feature decoupling and hazard absolute temperature rise. When the program is executed by a processor, it implements the steps of the method for automatic identification of power inspection infrared feature decoupling and hazard absolute temperature rise as described in any one of claims 1 to 7.

Images