Substation equipment temperature measurement method and system based on visible light guidance

Abstract

The invention relates to the technical field of mode recognition, in particular to a transformer substation equipment temperature measurement method and system based on visible light guidance. Firstly, a visible light image, an infrared thermogram and vibration spectrum data are synchronously collected, a dynamic compensation coefficient matrix is constructed through illumination intensity analysis, the dynamic compensation coefficient matrix is applied to the infrared thermogram, and illumination interference compensation is achieved. And then, inputting the corrected infrared thermogram and visible light image into a pre-trained heterogeneous atlas matching model, and identifying the category and three-dimensional coordinates of the target equipment. And based on the equipment category information, calling a standard temperature field template in the digital twin model library, establishing a heat conduction path model in combination with the vibration spectrum data, and generating a temperature field prediction reference map of the equipment. And finally, through pixel-by-pixel deviation analysis, inputting the temperature deviation value, the image features and the vibration data into the probability graphic model, and outputting equipment fault information. The method can realize high-precision temperature measurement and fault diagnosis.

Description

Technical Field

[0001] The present invention relates to the field of pattern recognition, and specifically to a method and system for temperature measurement of substation equipment based on visible light guidance. Background Technique

[0002] With the development of smart grids and digital substations, the real-time monitoring and intelligent diagnosis of equipment operating status have become important links to ensure the safe and stable operation of power systems. During the long-term operation of substation equipment, local heating often occurs due to factors such as aging, poor contact, and overload, which may further lead to equipment failures and even safety accidents. Therefore, accurate and timely temperature monitoring and fault warning of key substation equipment are of great significance.

[0003] In the prior art, infrared thermal imaging technology has been widely used for non-contact temperature measurement of substation equipment. However, limited by problems such as environmental light interference, insufficient target recognition accuracy, and low infrared image resolution, there are often defects such as large temperature measurement errors and high false alarm rates. Especially under complex background conditions, it is difficult to accurately identify equipment types and perform fine temperature measurement only relying on infrared images. In recent years, multi-modal perception methods that fuse visible light and infrared information have gradually attracted attention. By using visible light images to assist in correcting, positioning, and identifying infrared images, the temperature measurement accuracy and robustness can be improved to a certain extent. However, existing solutions are mostly limited to image-level registration and temperature threshold judgment, lacking in-depth modeling of equipment operating conditions and state prediction mechanisms.

[0004] Therefore, a method and system for temperature measurement of substation equipment based on visible light guidance are proposed. Summary of the Invention

[0005] The present invention provides a method and system for temperature measurement of substation equipment based on visible light guidance. By synchronously collecting visible light images, infrared thermal images, and vibration spectrum data, performing light compensation on the visible light images, and correcting the infrared thermal images; using a heterogeneous graph matching model to identify the equipment type and its three-dimensional coordinates, and generating a temperature field prediction reference map in combination with a digital twin model; by comparing the deviation between the real-time temperature and the predicted temperature, and combining the vibration data, using a probabilistic graphical model to output equipment fault information.

[0006] To achieve the above object, the present invention provides the following technical solutions.

[0007] A method for temperature measurement of substation equipment based on visible light guidance includes: Synchronously collecting multi-modal data of target substation equipment, where the multi-modal data includes visible light images, infrared thermal images, and vibration spectrum data; Performing an analysis on the light intensity distribution of the visible light image, constructing a dynamic compensation coefficient matrix based on the analysis results, and applying the compensation coefficient matrix to the infrared thermal image to generate a corrected infrared thermal image; Input the corrected infrared thermal image and visible light image into a pre-trained heterogeneous graph spectral matching model to output the category information of the target device and the three-dimensional spatial coordinates of the target device in the image space; Based on the category information of the target device, retrieve the corresponding standard temperature field template from the digital twin model library, and combine the vibration spectrum data to establish a heat conduction path model to generate a temperature field prediction reference map for the target device in the current state; Perform pixel-by-pixel deviation analysis on the corrected infrared thermal image and the temperature field prediction reference map to calculate the temperature deviation value between the real-time temperature field and the predicted temperature field; Input the temperature deviation value, image feature information, and vibration spectrum data as observation variables into a probabilistic graphical model to output the fault information of the target device.

[0008] Further, the steps of generating the corrected infrared thermal image include: Perform gray-scale transformation and illumination mean filtering on the visible light image to obtain an image illumination intensity distribution map; Construct a dynamic compensation coefficient matrix based on the illumination intensity distribution map, where the compensation coefficient of each pixel point is determined according to the deviation between the illumination brightness at the corresponding position of the pixel point and the set reference brightness; Apply the dynamic compensation coefficient matrix to the original temperature pixel values of the infrared thermal image and perform illumination interference compensation through pixel-level point multiplication to obtain the corrected infrared thermal image.

[0009] Further, the heterogeneous graph spectral matching model includes: A device recognition unit for combining the visible light image and the corrected infrared thermal image to identify the category information of the target device through a convolutional neural classification model; An image feature extraction unit for extracting key points and their corresponding descriptors from the visible light image and the corrected infrared thermal image through the SuperPoint model; A feature matching unit for performing feature matching on the key point descriptors through the Euclidean distance and eliminating incorrect matches through cross-validation; A spatial geometry estimation unit for outputting the three-dimensional spatial coordinates of the target device in the image space based on a geometric model fitting method of the random sample consensus algorithm.

[0010] Further, the steps of generating the temperature field prediction reference map include: According to the category information of the target device, retrieve a standard temperature field template that matches the structure, material, and working parameters of the target device from the digital twin model library; Extracting the operating state parameters of the target device using the vibration spectrum data, including frequency peak, spectrum width and spectrum line energy distribution, and inputting the operating state parameters into the device heat conduction simulation module to calculate the temperature conduction path of the device in the current operating state by using the finite element method; The temperature field is reconstructed and mapped based on the temperature conduction path and the standard temperature field template to generate a temperature field prediction reference map suitable for the current working state.

[0011] Furthermore, the step of outputting the current device fault information includes: The temperature deviation value, image feature information and vibration spectrum data are used as observation variables to construct an observation vector and input it into a predefined probability graphical model to perform joint probability inference on the health status of the target device; The maximum a posteriori estimation method is used to solve the occurrence probability of various typical failure modes and output the probability distribution vector of the target equipment under various failure types; Fault information including a fault risk level and a most likely fault type is generated according to the probability distribution vector.

[0012] The present invention also provides a substation equipment temperature measurement system based on visible light guidance, comprising: Data acquisition module, used to synchronously collect multimodal data of target substation equipment, including visible light images, infrared thermal images and vibration spectrum data; The infrared thermal image correction module is used to analyze the light intensity distribution of the visible light image, construct a dynamic compensation coefficient matrix based on the analysis results, and apply the compensation coefficient matrix to the infrared thermal image to generate a corrected infrared thermal image; An information acquisition module is used to input the corrected infrared thermal image and the visible light image into a pre-trained heterogeneous spectrum matching model, and output the category information of the target device and the three-dimensional spatial coordinates of the target device in the image space; The simulation model building module is used to retrieve the corresponding standard temperature field template from the digital twin model library based on the category information of the target device, and to establish a heat conduction path model in combination with the vibration spectrum data to generate a temperature field prediction benchmark map of the target device in the current state; The temperature difference analysis module is used to perform pixel-by-pixel deviation analysis on the corrected infrared thermal image and the temperature field prediction reference image, and calculate the temperature deviation value between the real-time temperature field and the predicted temperature field; The faulty equipment acquisition module is used to input the temperature deviation value, image feature information, and vibration spectrum data as observation variables into the probabilistic graphical model and output the fault information of the target equipment.

[0013] Furthermore, the step of generating the corrected infrared thermal image includes: Perform gray-scale transformation and illumination mean filtering on the visible light image to obtain an image illumination intensity distribution map; Construct a dynamic compensation coefficient matrix based on the illumination intensity distribution map, where the compensation coefficient of each pixel is determined according to the deviation between the illumination brightness at the corresponding position of the pixel and the set reference brightness; Apply the dynamic compensation coefficient matrix to the original temperature pixel values of the infrared thermal image, and perform illumination interference compensation through pixel-level point multiplication to obtain the corrected infrared thermal image.

[0014] Furthermore, the heterogeneous graph matching model includes: A device recognition unit, which is used to combine the visible light image and the corrected infrared thermal image, and identify the category information of the target device through a convolutional neural classification model; An image feature extraction unit, which is used to extract key points and their corresponding descriptors from the visible light image and the corrected infrared thermal image through the SuperPoint model; A feature matching unit, which is used to perform feature matching on the key point descriptors through the Euclidean distance, and eliminate incorrect matches through cross-validation; A spatial geometry estimation unit, which is used to output the three-dimensional spatial coordinates of the target device in the image space based on the geometric model fitting method of the random sample consensus algorithm.

[0015] Furthermore, the steps of generating the temperature field prediction reference map include: According to the category information of the target device, retrieve a standard temperature field template that matches the structure, material, and working parameters of the target device from the digital twin model library; Use the vibration spectrum data to extract the operating state parameters of the target device, including frequency peak value, spectrum width, and spectral line energy distribution, and input the operating state parameters into the device heat conduction simulation module to calculate the temperature conduction path of the device under the current operating state through the finite element method; Perform temperature field reconstruction and mapping based on the temperature conduction path and the standard temperature field template to generate a temperature field prediction reference map suitable for the current working state.

[0016] Furthermore, the steps of outputting the current device fault information include: Use the temperature deviation value, image feature information, and vibration spectrum data as observation variables to construct an observation vector and input it into a predefined probabilistic graphical model to perform joint probability inference on the health state of the target device; Solve the occurrence probabilities of various typical fault modes through the maximum a posteriori estimation method, and output the probability distribution vector of the target device under multiple fault types; Generate fault information including fault risk level and the most likely fault type according to the probability distribution vector.

[0017] The beneficial effects of the present invention are as follows: 1. By analyzing the illumination intensity of visible light images, constructing a dynamic compensation coefficient matrix and applying it to infrared thermal images, the interference of environmental illumination changes on infrared imaging results is effectively reduced, the accuracy and consistency of temperature images are improved, and more reliable basic data is provided for subsequent image fusion and equipment status evaluation.

[0018] 2. By combining the category information of the target device and vibration spectrum data, constructing a personalized heat conduction path model, and generating a prediction benchmark map based on the standard temperature field template in the digital twin model library, not only the accuracy and pertinence of temperature distribution prediction are improved, but also the dynamic perception ability of the system for the equipment operation state is enhanced, providing a scientific basis for fault diagnosis and temperature anomaly detection.

[0019] 3. Taking the temperature deviation value, image feature information and vibration spectrum data as multi-source observation variables and inputting them into the probabilistic graphical model, it is possible to realize the joint probability inference analysis of the equipment fault state on the basis of fusing multi-modal perception information, effectively improving the accuracy and robustness of fault recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, but do not constitute a limitation to the present invention. In the drawings: Figure 1 is a flowchart of a method for measuring the temperature of substation equipment based on visible light guidance provided by the present invention; Figure 2 is a structural diagram of a system for measuring the temperature of substation equipment based on visible light guidance provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] The following describes the preferred embodiments of the present invention with reference to the drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0022] Embodiment 1 A method for measuring the temperature of substation equipment based on visible light guidance, as Figure 1 shown, includes: S100: Synchronously collect multi-modal data of the target substation equipment, and the multi-modal data includes visible light images, infrared thermal images and vibration spectrum data; Specifically, within the target substation area, high-resolution industrial cameras or intelligent video cameras are used to obtain device appearance images and visible light information, and an infrared thermal imager is used to capture thermal images of the surface temperature distribution of the device in real time; a three-axis MEMS accelerometer is used to collect vibration spectrum data during device operation; to achieve time consistency between data sources, a GPS time synchronization module is used to perform unified clock synchronization on each acquisition module to ensure that the captured visible light images, thermal images, and vibration data match in the time dimension, which is conducive to subsequent fusion processing.

[0023] S200: Analyze the illumination intensity distribution of the visible light image, construct a dynamic compensation coefficient matrix based on the analysis results, and apply this compensation coefficient matrix to the infrared thermal image to generate a corrected infrared thermal image; Further, the steps of generating the corrected infrared thermal image include: Perform gray-scale transformation and illumination mean filtering on the visible light image to obtain an image illumination intensity distribution map; Construct a dynamic compensation coefficient matrix based on the illumination intensity distribution map, where the compensation coefficient of each pixel is determined according to the deviation between the illumination brightness at its corresponding position and the set reference brightness; Apply the dynamic compensation coefficient matrix to the original temperature pixel values of the infrared thermal image, and perform illumination interference compensation through pixel-level point multiplication to obtain the corrected infrared thermal image.

[0024] Specifically, obtain the visible light image corresponding to the space of the infrared thermal image and perform gray-scale processing on it to remove color information interference. Preferably, a standard gray-scale transformation algorithm (such as the weighted average method) is used to obtain a gray-scale image , apply Gaussian blur filtering to the gray-scale image to obtain an image illumination intensity distribution map , which is used to reflect the illumination brightness of each position in the image; for each pixel point in the visible light image, according to its local illumination value and the reference brightness the deviation between them is used to calculate the dynamic compensation coefficient matrix , where the reference brightness is the overall mean value of the image, and the calculation formula of the dynamic compensation coefficient matrix is: ; Among them, represents the pixel point coordinates, represents the illumination compensation adjustment factor, and the range is generally taken as [0.2, 1.0], and the value in this embodiment is 0.75; for each temperature pixel value in the infrared thermal image, perform a pixel-by-pixel multiplication operation with the corresponding element of the compensation coefficient matrix to obtain the corrected temperature value: ; Among them, each temperature pixel value in the infrared thermal image is corrected, and the generated image is the corrected infrared thermal image, which is used to reflect the true temperature distribution state after removing the influence of visible light illumination interference.

[0025] By analyzing the illumination intensity distribution of the visible light image, constructing a dynamic compensation coefficient matrix and applying it to the infrared thermal image, the influence of ambient light changes on the infrared temperature measurement accuracy is effectively reduced. Compared with the traditional static calibration or single filtering method, this method can achieve pixel-level adaptive correction, improve the temperature expression accuracy and spatial consistency of the infrared image, and provide more stable and reliable basic data support for subsequent equipment identification and temperature field analysis.

[0026] S300: Input the corrected infrared thermal image and the visible light image into a pre-trained heterogeneous graph matching model, and output the category information of the target device and the three-dimensional spatial coordinates of the target device in the image space; Furthermore, the heterogeneous graph matching model includes: An equipment identification unit, which is used to combine the visible light image and the corrected infrared thermal image, and identify the category information of the target device through a convolutional neural classification model; An image feature extraction unit, which is used to extract key points and their corresponding descriptors from the visible light image and the corrected infrared thermal image through the SuperPoint model; A feature matching unit, which is used to perform feature matching on the key point descriptors through the Euclidean distance, and remove incorrect matches through the cross-validation method; A spatial geometry estimation unit, which is used to output the three-dimensional spatial coordinates of the target device in the image space based on the geometric model fitting method of the random sample consensus algorithm.

[0027] Specifically, the corrected infrared thermal image and the corresponding visible light image are used as inputs. The category information of the target device is identified through a convolutional neural classification model. Here, the convolutional neural classification model is not limited. It can be a single-channel fusion network that fuses the infrared thermal image and the visible light image, or a two-channel model that separately inputs the infrared thermal image and the visible light image into different channels. In this embodiment, a two-channel Daul-YOLO model is preferably used; the SuperPoint model is used to extract key points and calculate descriptors for the two images respectively. SuperPoint is a lightweight neural network architecture that includes a shared encoder and two branch decoders, which generate a key point heat map and a descriptor tensor respectively; based on the Euclidean distance metric, the descriptors extracted by SuperPoint are pairwise matched, and the cross-validation method is used to eliminate the mismatched pairs, and only the key point pairs that are the nearest neighbors to each other are retained as matching points to enhance the accuracy of the matching; the random sample consensus algorithm is used to fit the spatial geometric model for the matching point pairs, and the three-dimensional spatial coordinates of the target device in the two-dimensional image space are further calculated according to the image shooting parameters.

[0028] By constructing a heterogeneous graph spectral matching model and fusing the information of the visible light image and the corrected infrared thermal image, both the accurate identification of the device category and the three-dimensional spatial coordinate positioning of the device in the image space are realized. Among them, the SuperPoint key point extraction and matching mechanism improves the feature consistency between multi-modal images. Combined with the geometric fitting method of the RANSAC algorithm, the matching robustness and positioning accuracy are significantly enhanced. This method effectively improves the automation level of device identification and positioning and the spatial accuracy of the temperature measurement task, laying a high-quality data foundation for subsequent thermal field analysis and fault diagnosis.

[0029] S400: Based on the category information of the target device, retrieve the corresponding standard temperature field template from the digital twin model library, and combine the vibration spectrum data to establish a heat conduction path model to generate a temperature field prediction benchmark map of the target device in the current state; Further, the steps of generating the temperature field prediction benchmark map include: According to the category information of the target device, retrieve the standard temperature field template that matches the structure, material, and working parameters of the device from the digital twin model library; Use the vibration spectrum data to extract the operating state parameters of the device, including frequency peak, spectrum width, and spectral line energy distribution, and input the operating state parameters into the device heat conduction simulation module to calculate the temperature conduction path of the device in the current operating state through the finite element method; Based on the temperature conduction path and the standard temperature field template, perform temperature field reconstruction and mapping to generate a temperature field prediction benchmark map adapted to the current working state.

[0030] Specifically, retrieve the standard temperature field template that matches the device from the pre-constructed digital twin model library. This template is constructed from the actual device structure, material properties (such as thermal conductivity, specific heat capacity), workload, and historical operating condition data. Conduct a frequency-domain analysis on the synchronously collected vibration spectrum data to extract the key operating state parameters of the current device, specifically including the frequency peak, spectrum width, and spectral line energy distribution. Use the above operating state parameters as inputs and pass them into the device heat conduction simulation module. This module establishes a heat conduction path model of the device based on the finite element method (FEM), including the thermal coupling relationship of the internal structure, external heat exchange boundary conditions, and temperature evolution over time. Integrate the heat conduction path diagram with the standard temperature field template. During the generation of the temperature field prediction reference diagram, first, based on the heat conduction path model and the standard temperature field template in the digital twin model library, reconstruct the temperature field distribution diagram under the current operating state through the finite element simulation method to achieve temperature field image reconstruction. Subsequently, use the spatial projection registration and numerical amplitude matching method to map the reconstructed temperature field diagram to the infrared image coordinate system and calibrate and adjust the temperature amplitude to achieve the consistency of the prediction diagram and the infrared image in space and numerically, thereby generating a temperature field prediction reference diagram that matches the actual observation conditions. This prediction reference diagram reflects the theoretically expected temperature distribution of the device under the current operating state and is used for subsequent deviation analysis with the actual infrared thermal image.

[0031] The above technical solution can dynamically generate a temperature field prediction reference diagram that conforms to the actual operating state by introducing a digital twin model and a heat conduction simulation mechanism, combining the category information of the device and real-time vibration spectrum data. This method not only overcomes the limitation that traditional static templates cannot reflect the current heat distribution characteristics of the device but also accurately simulates the temperature conduction path through the finite element method, achieving a high-precision prediction of the device's temperature rise trend and improving the scientific nature of temperature measurement analysis and the accuracy of fault prediction.

[0032] S500: Conduct a pixel-by-pixel deviation analysis on the corrected infrared thermal image and the temperature field prediction reference diagram, and calculate the temperature deviation value between the real-time temperature field and the predicted temperature field. Specifically, to ensure the consistency of the actual infrared thermal image and the predicted temperature field reference diagram in the spatial dimension, first perform precise registration operations on the two images, including rigid registration based on image features (such as edges, corners, or key structure contours), and then resample the corrected infrared image using an affine transformation to ensure that the same pixel position corresponds to the same physical point of the device in both images. After completing the image alignment, use the pixel-by-pixel difference operation method to calculate the temperature deviation value. Let be the predicted temperature of the corresponding pixel in the temperature field prediction reference diagram, then the temperature deviation value of the corresponding pixel is: ; Among them, is the actual temperature of the pixels in the calibrated infrared thermal image obtained in step S200 , represents the temperature deviation value corresponding to the pixel point.

[0033] S600: Input the temperature deviation value, image feature information, and vibration spectrum data as observation variables into the probabilistic graphical model, and output the fault information of the target device.

[0034] Furthermore, the steps of outputting the current device fault information include: Construct an observation vector with the temperature deviation value, image feature information, and vibration spectrum data as observation variables and input it into a predefined probabilistic graphical model to perform joint probability inference on the health state of the target device; Solve the occurrence probabilities of various typical fault modes through the maximum a posteriori estimation method, and output the probability distribution vector of the current device under multiple fault types; Generate fault information including fault risk level and the most likely fault type according to the probability distribution vector.

[0035] Furthermore, unify the three main observation variables into a standard feature input format, including temperature deviation value, image feature information, and vibration spectrum data. The image feature information includes features such as texture information, edge density, and color gradient distribution from visible light images and calibrated infrared thermal images. This part of the features is obtained through the convolutional layer of the convolutional neural network. After standard normalization processing of these three main observation variables, they are used as input observation vectors for the next modeling process; Use a Bayesian network as the core probabilistic graphical model for fault causal reasoning. This model structure is pre-constructed based on expert knowledge and device fault historical data, and includes observation variable nodes (corresponding to input multi-modal features), hidden state nodes (representing device operating state parameters such as local overheating, insulation aging, etc.), and fault type nodes (outputting the most likely fault modes such as poor contact, overload operation, abnormal heat dissipation, etc.); Use the maximum a posteriori estimation method to solve the model and output the probability distribution vector of the device under multiple fault types; Through post-processing and decision extraction, convert the probability distribution vector into a fault risk level and the most likely fault type. Among them, the fault type determination takes the fault label corresponding to the maximum value index from the probability distribution vector. This step can be described as: ; wherein, represents the fault type , represents the fault type probability of, represents the maximization function, and the fault risk level is judged by dividing the interval with the maximum fault probability value and the threshold. For example, when, the risk level is extremely high risk.

[0036] By fusing the temperature deviation value, image feature information, and vibration spectrum data into an observation vector and introducing a probabilistic graphical model for joint probability inference, it is possible to comprehensively evaluate the operating state of the device under the collaboration of multi-source information. Using the maximum a posteriori estimation method to calculate the probability distribution of various fault modes helps to accurately identify the most likely fault type and evaluate its risk level, thereby providing a scientific basis for operation and maintenance decisions and improving the accuracy of fault diagnosis and the intelligent level of the system.

[0037] Embodiment 2 The present invention also provides a temperature measurement system for substation equipment based on visible light guidance, as Figure 2 shown, including: A data acquisition module for synchronously acquiring multi-modal data of target substation equipment, where the multi-modal data includes visible light images, infrared thermal images, and vibration spectrum data; An infrared thermal image correction module for analyzing the illumination intensity distribution of visible light images, constructing a dynamic compensation coefficient matrix based on the analysis results, and applying the compensation coefficient matrix to infrared thermal images to generate corrected infrared thermal images; An information acquisition module for inputting the corrected infrared thermal images and visible light images into a pre-trained heterogeneous graph spectral matching model and outputting the category information of the target device and the three-dimensional spatial coordinates of the target device in the image space; A simulation model establishment module for retrieving the corresponding standard temperature field template from the digital twin model library based on the category information of the target device and establishing a heat conduction path model in combination with vibration spectrum data to generate a temperature field prediction reference map of the target device in the current state; A temperature difference analysis module for performing pixel-by-pixel deviation analysis on the corrected infrared thermal images and the temperature field prediction reference map to calculate the temperature deviation value between the real-time temperature field and the predicted temperature field; A faulty device acquisition module for inputting the temperature deviation value, image feature information, and vibration spectrum data as observation variables into a probabilistic graphical model and outputting the fault information of the target device.

[0038] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A temperature measurement method for substation equipment based on visible light guidance, characterized in that: include: Synchronously collect multimodal data of target substation equipment, including visible light images, infrared thermal images and vibration spectrum data; Performing light intensity distribution analysis on the visible light image, constructing a dynamic compensation coefficient matrix based on the analysis results, and applying the compensation coefficient matrix to the infrared thermal image to generate a corrected infrared thermal image; Input the corrected infrared thermal image and visible light image into the pre-trained heterogeneous graph matching model, and output the category information of the target device and the three-dimensional spatial coordinates of the target device in the image space; Based on the category information of the target device, the corresponding standard temperature field template is retrieved from the digital twin model library, and combined with the vibration spectrum data, a heat conduction path model is established to generate a temperature field prediction benchmark map of the target device in the current state; Perform pixel-by-pixel deviation analysis on the corrected infrared thermal image and the temperature field prediction reference image to calculate the temperature deviation value between the real-time temperature field and the predicted temperature field; The temperature deviation value, image feature information, and vibration spectrum data are input into the probability graphical model as observation variables, and the fault information of the target equipment is output.

2. A method for measuring temperature of substation equipment based on visible light guidance according to claim 1, characterized in that: The step of generating the corrected infrared thermal image comprises: Performing grayscale transformation and illumination mean filtering on the visible light image to obtain an image illumination intensity distribution map; Constructing a dynamic compensation coefficient matrix based on the light intensity distribution diagram, wherein the compensation coefficient of each pixel is determined according to the deviation between the light brightness at the corresponding position of the pixel and the set reference brightness; The dynamic compensation coefficient matrix is ​​applied to the original temperature pixel value of the infrared thermal image, and the light interference compensation is performed by pixel-level dot multiplication to obtain the corrected infrared thermal image.

3. According to a method for measuring temperature of substation equipment based on visible light guidance according to claim 1, it is characterized in that: The heterogeneous graph matching model includes: A device identification unit is used to combine the visible light image and the corrected infrared thermal image to identify the category information of the target device through a convolutional neural classification model; An image feature extraction unit, used to extract key points and their corresponding descriptors from visible light images and corrected infrared thermal images through a SuperPoint model; A feature matching unit is used to perform feature matching on key point descriptors by using Euclidean distance and eliminate false matches by cross-validation; The spatial geometry estimation unit is used for outputting the three-dimensional spatial coordinates of the target device in the image space based on the geometric model fitting method based on the random sampling consistency algorithm.

4. The method for measuring temperature of substation equipment based on visible light guidance according to claim 1 is characterized in that: The step of generating the temperature field prediction reference map comprises: According to the category information of the target device, a standard temperature field template matching the structure, material and working parameters of the target device is retrieved from the digital twin model library; Extracting the operating state parameters of the target device using the vibration spectrum data, including frequency peak, spectrum width and spectrum line energy distribution, and inputting the operating state parameters into the device heat conduction simulation module to calculate the temperature conduction path of the device in the current operating state by the finite element method; The temperature field is reconstructed and mapped based on the temperature conduction path and the standard temperature field template to generate a temperature field prediction reference map suitable for the current working state.

5. The method for measuring temperature of substation equipment based on visible light guidance according to claim 1 is characterized in that: The steps for outputting the current device fault information include: The temperature deviation value, image feature information and vibration spectrum data are used as observation variables to construct an observation vector and input it into a predefined probability graphical model to perform joint probability inference on the health status of the target device; The probability of occurrence of various typical failure modes is solved by the maximum a posteriori estimation method, and the probability distribution vector of the target equipment under various failure types is output; Fault information including a fault risk level and a most likely fault type is generated according to the probability distribution vector.

6. A temperature measurement system for substation equipment based on visible light guidance, characterized in that: include: Data acquisition module, used to synchronously collect multimodal data of target substation equipment, including visible light images, infrared thermal images and vibration spectrum data; The infrared thermal image correction module is used to analyze the light intensity distribution of the visible light image, construct a dynamic compensation coefficient matrix based on the analysis results, and apply the compensation coefficient matrix to the infrared thermal image to generate a corrected infrared thermal image; An information acquisition module is used to input the corrected infrared thermal image and the visible light image into a pre-trained heterogeneous spectrum matching model, and output the category information of the target device and the three-dimensional spatial coordinates of the target device in the image space; The simulation model building module is used to retrieve the corresponding standard temperature field template from the digital twin model library based on the category information of the target device, and to establish a heat conduction path model in combination with the vibration spectrum data to generate a temperature field prediction benchmark map of the target device in the current state; The temperature difference analysis module is used to perform pixel-by-pixel deviation analysis on the corrected infrared thermal image and the temperature field prediction reference image, and calculate the temperature deviation value between the real-time temperature field and the predicted temperature field; The faulty equipment acquisition module is used to input the temperature deviation value, image feature information, and vibration spectrum data as observation variables into the probabilistic graphical model and output the fault information of the target equipment.

7. A temperature measurement system for substation equipment based on visible light guidance according to claim 6, characterized in that: The step of generating the corrected infrared thermal image comprises: Performing grayscale transformation and illumination mean filtering on the visible light image to obtain an image illumination intensity distribution map; Constructing a dynamic compensation coefficient matrix based on the light intensity distribution diagram, wherein the compensation coefficient of each pixel is determined according to the deviation between the light brightness at the corresponding position of the pixel and the set reference brightness; The dynamic compensation coefficient matrix is ​​applied to the original temperature pixel value of the infrared thermal image, and the light interference compensation is performed by pixel-level dot multiplication to obtain the corrected infrared thermal image.

8. The temperature measurement system for substation equipment based on visible light guidance according to claim 6 is characterized in that: The heterogeneous graph matching model includes: A device identification unit is used to combine the visible light image and the corrected infrared thermal image to identify the category information of the target device through a convolutional neural classification model; An image feature extraction unit, used to extract key points and their corresponding descriptors from visible light images and corrected infrared thermal images through a SuperPoint model; A feature matching unit is used to perform feature matching on key point descriptors by using Euclidean distance and eliminate false matches by cross-validation; The spatial geometry estimation unit is used for outputting the three-dimensional spatial coordinates of the target device in the image space based on the geometric model fitting method based on the random sampling consistency algorithm.

9. The substation equipment temperature measurement system based on visible light guidance according to claim 6 is characterized in that: The step of generating the temperature field prediction reference map comprises: According to the category information of the target device, a standard temperature field template matching the structure, material and working parameters of the target device is retrieved from the digital twin model library; Extracting the operating state parameters of the target device using the vibration spectrum data, including frequency peak, spectrum width and spectrum line energy distribution, and inputting the operating state parameters into the device heat conduction simulation module to calculate the temperature conduction path of the device in the current operating state by the finite element method; The temperature field is reconstructed and mapped based on the temperature conduction path and the standard temperature field template to generate a temperature field prediction reference map suitable for the current working state.

10. A temperature measurement system for substation equipment based on visible light guidance according to claim 6, characterized in that: The steps for outputting the current device fault information include: The temperature deviation value, image feature information and vibration spectrum data are used as observation variables to construct an observation vector and input it into a predefined probability graphical model to perform joint probability inference on the health status of the target device; The probability of occurrence of various typical failure modes is solved by the maximum a posteriori estimation method, and the probability distribution vector of the target equipment under various failure types is output; Fault information including a fault risk level and a most likely fault type is generated according to the probability distribution vector.

Images