Equipment health assessment method, system and equipment in dynamic operation and maintenance of power distribution network based on point cloud three-dimensional visualization and storage medium

Abstract

The invention discloses an equipment health assessment method, system and equipment in dynamic operation and maintenance of a power distribution network based on point cloud three-dimensional visualization and a storage medium, and relates to the technical field of power distribution network monitoring and assessment, and the method comprises the steps: obtaining the original point cloud data of a power distribution network scene and the operation state data of target equipment in the power distribution network; processing the original point cloud data to construct a three-dimensional point cloud model of the power distribution network scene, and performing semantic segmentation on the equipment point cloud; based on the historical operation state data of the target equipment, performing time sequence prediction on the degradation index of the operation health state of the target equipment by adopting a GPR (Gaussian Process Regression) model to obtain a probabilistic prediction result containing a prediction mean value and a prediction variance; evaluating the current comprehensive health state of the target equipment by adopting a fuzzy analytic hierarchy process (FAHP) model to obtain a comprehensive health evaluation score; dynamically adjusting the fusion weight of the probabilistic prediction result and the comprehensive health assessment score by using the prediction variance, and calculating to obtain a hybrid prediction health index HPHI; and respectively mapping the HPHI and the prediction variance into visual attributes, and rendering the target equipment in the three-dimensional scene. According to the method, the dynamic visual accurate quantitative evaluation of the health state of the target equipment and the reliable prediction of the future trend in the dynamic operation and maintenance of the power distribution network are realized.

Description

Technical Field

[0001] This invention relates to the field of distribution network monitoring and evaluation technology, and in particular to a distribution network monitoring and evaluation system based on point cloud 3D visualization. Methods, systems, equipment, and storage media for equipment health assessment in dynamic power grid operation and maintenance. Background Technology

[0002] As a key component of the power system, the efficiency and reliability of the operation and maintenance (O&M) work of the distribution network directly affect end users. However, existing technologies face a series of fundamental and interconnected technical bottlenecks in achieving intelligent and refined management of the distribution network, resulting in the long-term stagnation of the O&M model at an inefficient and high-risk level.

[0003] First, traditional operation and maintenance methods are inherently limited by the physical and cognitive limitations of manual inspection. Inspection methods relying on manual foot patrols and visual inspections are not only labor-intensive and inefficient, but also pose extremely high safety risks in complex terrain. More importantly, the quality of the data produced is inconsistent, consisting mostly of qualitative descriptions and non-standardized images, lacking the precision and dimensions necessary to support advanced data analysis and accurate decision-making, fundamentally limiting the level of precision in operation and maintenance.

[0004] Secondly, existing digital tools have failed to solve the fundamental problem of data fusion, creating "digital silos." Although tools such as Geographic Information Systems (GIS) and Supervisory Control and Data Acquisition (SCADA) systems have been introduced, they operate independently. GIS systems lack true three-dimensional geometric information of the physical world, while SCADA systems lack spatial context. A significant gap exists between the two, making it impossible to effectively integrate precise three-dimensional spatial geometry with real-time dynamic operational data. This results in managers being unable to obtain a comprehensive and intuitive situational awareness of the power grid asset status.

[0005] Furthermore, while emerging remote sensing technologies solve data acquisition problems, they also create more severe data processing bottlenecks. The application of UAVs and LiDAR technologies, while overcoming the limitations of manual data acquisition, generates terabytes of highly unstructured 3D point cloud data. Existing data processing technologies, whether traditional geometric algorithms or early machine learning methods, struggle to efficiently process such massive datasets, often requiring complex preprocessing and being unsuitable for large-scale scenarios. This bottleneck in data processing efficiency has become the biggest obstacle to transforming high-precision 3D data into usable digital assets, hindering the construction of high-fidelity digital twins.

[0006] Ultimately, these bottlenecks collectively result in operational strategies remaining in a passive "post-event response" mode. Due to the lack of a unified platform capable of integrating multi-source data, processing it efficiently, and presenting it intuitively, current operational work remains centered on fault repair and periodic maintenance. Even with the deployment of condition monitoring sensors, the lack of an advanced analysis engine that integrates with 3D physical models prevents accurate quantitative assessment of equipment health status and reliable prediction of future trends, thus hindering the implementation of true predictive maintenance (PdM). Summary of the Invention

[0007] Therefore, the technical problem to be solved by this invention is to solve the problem that in the dynamic operation and maintenance of power distribution networks, it is impossible to achieve accurate quantitative assessment of equipment health status and reliable prediction of future trends, and the data cannot be presented in a dynamic and visual manner.

[0008] The above-mentioned technical problems are solved by the following technical solution: This invention proposes a point cloud-based 3D... A method, system, device, and storage medium for visualizing equipment health assessment in dynamic operation and maintenance of distribution networks, comprising a point cloud-based 3D visualization method for equipment health assessment in dynamic operation and maintenance of distribution networks, a point cloud-based 3D visualization system for equipment health assessment in dynamic operation and maintenance of distribution networks, a computer device, and a computer-readable storage medium.

[0009] Firstly, this invention proposes a method for equipment health assessment in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization, which includes: Acquire raw point cloud data of the power distribution network scenario and operating status data of target equipment in the power distribution network; The raw point cloud data is processed to construct a 3D point cloud model of the power distribution network scenario, and semantic segmentation is performed on the device point cloud in the 3D point cloud model to identify and locate the target device. Based on the historical operating status data of the target equipment, the Gaussian process regression (GPR) model is used to make time series predictions on the degradation index that characterizes the operating health status of the target equipment, and the probabilistic prediction results including the prediction mean and prediction variance are obtained. Based on a pre-set health assessment index system, the fuzzy hierarchical analysis method (FAHP) model is used to assess the current comprehensive health status of the target equipment and obtain a comprehensive health assessment score. A hybrid predictive health index (HPHI) is calculated by using a pre-defined mathematical fusion algorithm and by dynamically adjusting the fusion weights of the probabilistic prediction results and the comprehensive health assessment score using the prediction variance. HPHI is mapped to a first visual attribute based on color gradient, and prediction variance is mapped to a second visual attribute to characterize the degree of prediction uncertainty. The second visual attribute is visually distinct from the first visual attribute. In the 3D point cloud model, the first and second visual attributes are rendered in real time on the point cloud corresponding to the target device to achieve dynamic visualization of the health status of the target device.

[0010] In a preferred embodiment of the equipment health assessment method for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization described in this invention: the GPR model adopts a composite kernel function. The composite kernel function is a linear combination of at least two basic kernel functions.

[0011] In a preferred embodiment of the equipment health assessment method for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization described in this invention: the composite kernel function The mathematical expression for is as follows: in, It is a quadratic exponential kernel function used to characterize the long-term smooth degradation trend of equipment; It is an exponential sine square kernel function, used to characterize periodic fluctuations caused by environmental factors; It is a rational quadratic kernel function used to characterize multi-scale perturbations caused by sudden operational events.

[0012] In a preferred embodiment of the equipment health assessment method for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization described in this invention: the fuzzy hierarchical analysis (FAHP) model is used to assess the current comprehensive health status of the target equipment. The specific steps are as follows: Construct a multi-level health assessment indicator system that includes a target layer, a criterion layer, and an indicator layer; The triangular fuzzy number TFN is used to compare the relative importance of indicators in each level pairwise to construct a fuzzy pairwise comparison matrix. Calculate the fuzzy weights of each indicator, and obtain the deterministic weights of each indicator through the defuzzification method; The comprehensive health assessment score is obtained by weighting the normalized measured values ​​of each indicator and their deterministic weights.

[0013] In a preferred embodiment of the equipment health assessment method for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization described in this invention: the specific steps for dynamically adjusting the fusion weight of probabilistic prediction results and comprehensive health assessment scores using prediction variance are as follows. Transform the probabilistic prediction result into a probabilistic risk score. ; Based on prediction variance Calculate a dynamic weighting coefficient Dynamic weighting coefficient The mathematical expression for is as follows: in, For the GPR model at future moments The prediction variance This is a preset sensitivity constant; Through dynamic weighting coefficients Score for probabilistic risk and comprehensive health assessment score Perform weighted fusion.

[0014] In a preferred embodiment of the device health assessment method for dynamic operation and maintenance of power distribution network based on point cloud 3D visualization described in this invention: the second visualization attribute is a dynamic halo around the target device, and the prediction variance is mapped to at least one visual feature of the dynamic halo, wherein the visual feature is selected from the radius, transparency or pulsation frequency of the halo.

[0015] Secondly, this invention proposes a device health assessment system for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization, comprising: The data acquisition and preprocessing module is used to acquire and process raw point cloud data and equipment operating status data of the power distribution network; The point cloud management module is used to build, store, and manage 3D point cloud models of power distribution network scenarios, and to perform semantic segmentation on the device point clouds in the model. The predictive modeling module is used to run a Gaussian process regression (GPR) model to perform time series predictions on the target equipment—a degradation index characterizing its operational health status—and output probabilistic prediction results including the predicted mean and predicted variance. The health assessment module is used to run a fuzzy analytic hierarchy process (FAHP) model to assess the current comprehensive health status of the target device and output a comprehensive health assessment score. The HPHI fusion engine module is used to calculate a hybrid predictive health index HPHI based on the outputs of the predictive modeling module and the health assessment module, and by dynamically adjusting the fusion weights using the predictive variance. The dynamic visualization module includes a GPU-accelerated rendering pipeline for rendering the visualization attributes corresponding to HPHI and prediction variance to the target device in the 3D point cloud model in real time.

[0016] In a preferred embodiment of the equipment health assessment system for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization described in this invention: the point cloud management module adopts an adaptive octree data structure to organize the 3D point cloud model. The leaf nodes or point cluster nodes of the octree contain pointers to a dynamic data block. The dynamic data block is used to store the time-series data, current HPHI value and prediction variance of the equipment associated with the node.

[0017] Thirdly, the present invention proposes a computer device, including a memory, a processor, and computer-executable instructions stored in the memory, wherein the processor is used to execute the computer-executable instructions to implement the steps of the equipment health assessment method in the dynamic operation and maintenance of the power distribution network based on point cloud 3D visualization.

[0018] Fourthly, the present invention proposes a computer-readable storage medium storing a program thereon, which, when executed by a processor, implements the steps of the equipment health assessment method for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization.

[0019] The beneficial effects of this invention are as follows: by processing the original point cloud data to construct a three-dimensional point cloud model of the power distribution network scenario and performing semantic segmentation on the equipment point cloud, the Gaussian process regression (GPR) model is used to predict the mean and variance of the target equipment's operational health status degradation index, and the fuzzy hierarchical analysis (FAHP) model is used to obtain the target equipment's comprehensive health assessment score. The HPHI is calculated by dynamically adjusting the fusion weights of the two using the predicted variance, and then the HPHI and predicted variance are mapped to visualization attributes and the target equipment is rendered in the three-dimensional scene. This achieves accurate quantitative assessment of the target equipment's health status and reliable prediction of future trends in the dynamic operation and maintenance of the power distribution network. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Obviously, the drawings described below only relate to some embodiments of the present invention and are not intended to limit the present invention. Wherein: Figure 1 A flowchart is shown for a method of equipment health assessment in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization; Figure 2 A schematic diagram of an equipment health assessment system for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization is shown. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0022] The terminology used in this invention is that which is currently widely used in the art in consideration of the function of the invention; however, these terms may vary according to the intent of those skilled in the art, precedent, or new technology in the art. Furthermore, specific terms may be chosen by the applicant, and in such cases, their detailed meanings will be described in the detailed description of the invention. Therefore, the terms used in this specification should not be construed as simple names, but rather based on their meanings and the overall description of the invention.

[0023] Reference Figure 1 This embodiment provides a method for equipment health assessment in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization, including the following steps: Acquire raw point cloud data of the power distribution network scenario and operating status data of target equipment in the power distribution network; The raw point cloud data is processed to construct a 3D point cloud model of the power distribution network scenario, and semantic segmentation is performed on the device point cloud in the 3D point cloud model to identify and locate the target device. Based on the historical operating status data of the target equipment, the Gaussian process regression (GPR) model is used to make time series predictions on the degradation index that characterizes the operating health status of the target equipment, and the probabilistic prediction results including the prediction mean and prediction variance are obtained. Based on a pre-set health assessment index system, the fuzzy hierarchical analysis method (FAHP) model is used to assess the current comprehensive health status of the target equipment and obtain a comprehensive health assessment score. A hybrid predictive health index (HPHI) is calculated by using a pre-defined mathematical fusion algorithm and by dynamically adjusting the fusion weights of the probabilistic prediction results and the comprehensive health assessment score using the prediction variance. HPHI is mapped to a first visual attribute based on color gradient, and prediction variance is mapped to a second visual attribute to characterize the degree of prediction uncertainty. The second visual attribute is visually distinct from the first visual attribute. In the 3D point cloud model, the first and second visual attributes are rendered in real time on the point cloud corresponding to the target device to achieve dynamic visualization of the health status of the target device.

[0024] Specifically, the process of the equipment health assessment method in the dynamic operation and maintenance of the distribution network based on point cloud 3D visualization is as follows: S1. Start the system and begin data acquisition and preprocessing; S2. Construct and update the global 3D point cloud model, and perform semantic segmentation; S3. The system enters the main loop and executes subsequent steps in parallel or serially for each target device in the monitoring list; S4. Based on the historical data of this device, perform GPR prediction and obtain the predicted mean. and prediction variance ; S5. Based on the current multi-source information of the device, perform a FAHP assessment to obtain a comprehensive health assessment score. ; S6. Combine the results of S4 and S5 to calculate the HPHI value and the uncertainty of the association; S7. Map the calculated HPHI and other information to visual attributes; S8. Render the updated device visualization in the 3D scene; S9. Determine whether monitoring needs to continue. If not, end the process. If yes, return to S3 to form a continuously updated dynamic operation and maintenance closed loop.

[0025] Specifically, the construction of a 3D point cloud model of the power distribution network scenario and the semantic segmentation of the device point cloud in the 3D point cloud model are as follows: Data acquisition and registration: High-precision laser scanning equipment is used to scan key scenes such as substations at multiple sites. The raw point cloud data is first filtered to remove noise points and outliers. Then, using the Iterative Closest Point (ICP) algorithm or a feature point-based registration algorithm, the point cloud data from different stations are precisely aligned to the same global coordinate system to form a complete and seamless scene point cloud. Semantic segmentation: In order for the system to "understand" the content of the point cloud model, semantic segmentation of the point cloud is required. Here, a supervised learning method is adopted. First, different categories of equipment (transformers, switch cabinets, insulators, etc.) and backgrounds (walls, ground, etc.) are manually labeled on the sample point cloud data. Then, a point cloud segmentation deep learning network (such as RandLA-Net) is trained using these labeled data. After training, the network can automatically classify new, unlabeled point cloud scenes at the pixel level and assign a semantic label to each point. Using an adaptive octree data structure: To efficiently manage and render massive point cloud data that may contain billions of points, an improved adaptiveoctree data structure is used. In this structure, the space is recursively divided into eight sub-cubes until the number of points contained in each leaf node is below a certain threshold or a preset maximum depth is reached. In addition to storing its spatial bounding box and statistical information of the point cloud data, each node also contains a dynamic data pointer that points to a dynamic data block independent of the octree geometry. This data block stores the device's time-series operating data, the latest HPHI value, prediction variance, and other attributes.

[0026] Specifically, the Gaussian process regression (GPR) model is used to perform time series prediction on the degradation index characterizing the operational health status of the target equipment, obtaining probabilistic prediction results including the predicted mean and predicted variance, as detailed below. GPR, as a nonparametric Bayesian regression method, assumes that the equipment degradation index... ,in Given training data, new time points can be predicted. The posterior distribution of the function values, with mean and variance, is as follows: Among them, prediction variance The uncertainty of the prediction is quantified. To accurately capture the complex degradation behavior of power equipment, the GPR model employs a composite kernel function. This composite kernel function is a linear combination of at least two basic kernel functions, and its mathematical expression is as follows: The definitions and physical meanings of each component are shown in the table below.

[0027] The set of hyperparameters for this composite kernel function is automatically optimized and learned by maximizing the marginal log-likelihood function of the observed data.

[0028] Specifically, the Fuzzy Hierarchical Analysis (FAHP) model is used to assess the current comprehensive health status of the target equipment, resulting in a comprehensive health assessment score, as detailed below. Construct a three-tiered evaluation system, including the objective layer, the criteria layer, and the indicator layer; Constructing a fuzzy comparison matrix: Invite domain experts to make pairwise comparisons based on their relative importance, and use triangular fuzzy numbers (TFN) to quantify these linguistic variables to construct a fuzzy pairwise comparison matrix; Calculating fuzzy weights and defuzzifying: The fuzzy weight vector of each indicator is calculated using the operational rules of fuzzy mathematics, and then the fuzzy weights are transformed into clear and deterministic weight values ​​using the defuzzification method; Weighted aggregation: The normalized real-time monitoring values ​​or assessment level scores of each underlying indicator are multiplied by their corresponding deterministic weights, and then aggregated layer by layer to obtain a comprehensive health assessment score. .

[0029] Specifically, the hybrid predictive health index HPHI is calculated by dynamically adjusting the fusion weights of the probabilistic prediction results and the comprehensive health assessment score using the prediction variance, as detailed below. Risk quantification of GPR prediction results: Transforming the probability distribution output by the GPR model into a dimensionless risk score. ; Derivation of dynamic weighting coefficient: A dynamic weighting coefficient is proposed, the value of which is related to the prediction uncertainty (i.e. prediction variance) of GPR; HPHI Final Fusion: Substituting the above results into the formula, we obtain the final HPHI calculation formula. This formula achieves dynamic fusion; when prediction certainty is high (small variance), HPHI is mainly determined by the current... The decision is that when forecast uncertainty is high (large variance), the huge uncertainty will also lower the overall health index through the weighting mechanism, which is in line with the intuition of risk management.

[0030] Specifically, HPHI is mapped to a first visual attribute based on a color gradient, and the prediction variance is mapped to a second visual attribute to characterize the degree of prediction uncertainty. The second visual attribute is visually distinct from the first visual attribute. In the 3D point cloud model, the first and second visual attributes are rendered in real time on the point cloud corresponding to the target device to achieve dynamic visualization of the health status of the target device, as detailed below. GPU-accelerated rendering pipeline: Employs a rendering pipeline based on modern graphics APIs (such as Vulkan) to fully utilize the parallel computing capabilities of the GPU, ensuring real-time rendering frame rates for large-scale point cloud scenes. LOD Management: When rendering each frame, based on the camera position and view frustum, the octree data structure is traversed to select the appropriate level of detail (LOD) for rendering, effectively controlling the amount of geometric data that needs to be processed for each frame; Mapping dynamic attributes to visual representation: In the fragment shader, obtain the device to which the current point belongs. The tuples are used to calculate the final pixel color according to preset mapping rules. For example, the HPHI value is mapped to a continuous color gradient from green to red, while the prediction variance is mapped to a dynamic halo around the device whose size or pulsation frequency varies with the variance.

[0031] The following uses a 35kV oil-immersed distribution transformer as an example to illustrate the implementation process of the method of the present invention.

[0032] Input point cloud data, time series data, and context data. Point cloud data: A high-precision point cloud model of the transformer and its station building was obtained through laser scanning; Time-series data: Obtain historical operating data of the transformer over the past two years, including hourly top oil temperature, load rate, and DGA data (especially acetylene C2H2 concentration). Contextual data: expert inspection records and historical fault data of equipment of the same model.

[0033] GPR calculation: C2H2 concentration is selected as the key degradation indicator. A GPR model with a composite kernel function is trained using time-series data from the past two years. The model predicts the C2H2 concentration one month in advance. Assuming at the current time t, the prediction for one month later is ( The concentration distribution of ) is .

[0034] FAHP Calculation: Based on a pre-defined hierarchical structure, experts perform pairwise comparisons of each criterion and indicator. Combining this with real-time monitoring data, these quantitative and qualitative information are input into the FAHP model to calculate the current comprehensive health assessment score. (In good condition).

[0035] HPHI calculation is performed through risk quantification and dynamic fusion. Risk quantification: Assuming a safe threshold for C2H2 concentration The value is 3 ppm. Calculate ; Dynamic fusion: Assuming sensitivity constant k=2, prediction variance ,calculate The output result is .

[0036] Visualization: In the 3D point cloud visualization interface, the point cloud of the transformer is rendered. The HPHI value is 0.8466, which is between green and yellow. Therefore, the transformer appears as a yellow-green color overall. Prediction variance... It is not a very small value. According to the mapping rules, a medium-sized, slowly pulsating, semi-transparent halo will appear around the transformer. When the maintenance personnel click on the transformer, a detailed information window will pop up on the screen, showing the current HPHI accurate value, the scores of each FAHP sub-item, and the future C2H2 concentration evolution curve predicted by GPR and its confidence interval. This provides a comprehensive, quantitative, and intuitive basis for decision-making in formulating the next monitoring or maintenance plan.

[0037] Reference Figure 2 This embodiment also provides a device health assessment system for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization, including: The data acquisition and preprocessing module is used to acquire and process raw point cloud data and equipment operating status data of the power distribution network; The point cloud management module is used to build, store, and manage 3D point cloud models of power distribution network scenarios, and to perform semantic segmentation on the device point clouds in the model. The predictive modeling module is used to run a Gaussian process regression (GPR) model to perform time series predictions on the target equipment—a degradation index characterizing its operational health status—and output probabilistic prediction results including the predicted mean and predicted variance. The health assessment module is used to run a fuzzy analytic hierarchy process (FAHP) model to assess the current comprehensive health status of the target device and output a comprehensive health assessment score. The HPHI fusion engine module is used to calculate a hybrid predictive health index HPHI based on the outputs of the predictive modeling module and the health assessment module, and by dynamically adjusting the fusion weights using the predictive variance. The dynamic visualization module includes a GPU-accelerated rendering pipeline for rendering the visualization attributes corresponding to HPHI and prediction variance to the target device in the 3D point cloud model in real time.

[0038] Specifically, during system operation, the modules work together to provide users with a closed-loop intelligent operation and maintenance platform. Users can freely roam, zoom, and query in three-dimensional space through the interactive interface provided by the system, and intuitively perceive the health status of the entire power distribution network.

[0039] This embodiment also provides a computer device applicable to the equipment health assessment method in dynamic operation and maintenance of distribution networks based on point cloud 3D visualization, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the equipment health assessment method in dynamic operation and maintenance of distribution networks based on point cloud 3D visualization as proposed in the above embodiment.

[0040] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0041] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the equipment health assessment method for dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0042] In summary, this invention achieves accurate quantitative assessment of the health status of target equipment and reliable prediction of future trends in dynamic visualization during the dynamic operation and maintenance of power distribution networks by: processing raw point cloud data to construct a 3D point cloud model of the power distribution network scenario and performing semantic segmentation on the equipment point cloud; using a Gaussian process regression (GPR) model to predict the mean and variance of the target equipment's operational health status degradation index; using a fuzzy hierarchical analysis (FAHP) model to obtain the target equipment's comprehensive health assessment score; dynamically adjusting the fusion weights of the two using the predicted variance to calculate the HPHI; and then mapping the HPHI and predicted variance into visual attributes and rendering the target equipment in the 3D scene.

[0043] Finally, it should be noted that the methods, systems, devices, and storage media described in detail above are merely embodiments, and those skilled in the art can modify these embodiments in different ways as long as they do not depart from the scope of the present invention.

Claims

1. A method for equipment health assessment in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization, characterized in that: include, Acquire raw point cloud data of the power distribution network scenario and operating status data of target equipment in the power distribution network; The raw point cloud data is processed to construct a 3D point cloud model of the power distribution network scenario, and semantic segmentation is performed on the device point cloud in the 3D point cloud model to identify and locate the target device. Based on the historical operating status data of the target equipment, the Gaussian process regression (GPR) model is used to make time series predictions on the degradation index that characterizes the operating health status of the target equipment, and the probabilistic prediction results including the prediction mean and prediction variance are obtained. Based on a pre-set health assessment index system, the fuzzy hierarchical analysis method (FAHP) model is used to assess the current comprehensive health status of the target equipment and obtain a comprehensive health assessment score. A hybrid predictive health index (HPHI) is calculated by using a pre-defined mathematical fusion algorithm and by dynamically adjusting the fusion weights of the probabilistic prediction results and the comprehensive health assessment score using the prediction variance. HPHI is mapped to a first visual attribute based on color gradient, and prediction variance is mapped to a second visual attribute to characterize the degree of prediction uncertainty. The second visual attribute is visually distinct from the first visual attribute. In the 3D point cloud model, the first and second visual attributes are rendered in real time on the point cloud corresponding to the target device to achieve dynamic visualization of the health status of the target device.

2. The method for equipment health assessment in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization according to claim 1, characterized in that: The GPR model employs a composite kernel function. The composite kernel function is a linear combination of at least two basic kernel functions.

3. The method for equipment health assessment in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization according to claim 2, characterized in that: The composite kernel function The mathematical expression for is as follows: in, It is a quadratic exponential kernel function used to characterize the long-term smooth degradation trend of equipment; It is an exponential sine square kernel function, used to characterize periodic fluctuations caused by environmental factors; It is a rational quadratic kernel function used to characterize multi-scale perturbations caused by sudden operational events.

4. The method for equipment health assessment in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization according to claim 3, characterized in that: The fuzzy hierarchical analysis (FAHP) model is used to assess the current overall health status of the target equipment. The specific steps are as follows: Construct a multi-level health assessment indicator system that includes a target layer, a criterion layer, and an indicator layer; The triangular fuzzy number TFN is used to compare the relative importance of indicators in each level pairwise to construct a fuzzy pairwise comparison matrix. Calculate the fuzzy weights of each indicator, and obtain the deterministic weights of each indicator through the defuzzification method; The comprehensive health assessment score is obtained by weighting the normalized measured values ​​of each indicator and their deterministic weights.

5. The method for equipment health assessment in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization according to claim 4, characterized in that: The specific steps for dynamically adjusting the fusion weights of probabilistic prediction results and comprehensive health assessment scores using prediction variance are as follows: Transform the probabilistic prediction result into a probabilistic risk score. ; Based on prediction variance Calculate a dynamic weighting coefficient Dynamic weighting coefficient The mathematical expression for is as follows: in, For the GPR model at future moments The prediction variance This is a preset sensitivity constant; Through dynamic weighting coefficients Score for probabilistic risk and comprehensive health assessment score Perform weighted fusion.

6. The method for equipment health assessment in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization according to claim 5, characterized in that: The second visualization attribute is a dynamic halo around the target device, and the prediction variance is mapped to at least one visual feature of the dynamic halo, which is selected from the radius, transparency, or pulsation frequency of the halo.

7. A system for assessing equipment health in dynamic operation and maintenance of a distribution network based on point cloud 3D visualization, wherein the system is based on the method for assessing equipment health in dynamic operation and maintenance of a distribution network based on point cloud 3D visualization as described in any one of claims 1-6, characterized in that: include, The data acquisition and preprocessing module is used to acquire and process raw point cloud data and equipment operating status data of the power distribution network; The point cloud management module is used to build, store, and manage 3D point cloud models of power distribution network scenarios, and to perform semantic segmentation on the device point clouds in the model. The predictive modeling module is used to run a Gaussian process regression (GPR) model to perform time series predictions on the target equipment—a degradation index characterizing its operational health status—and output probabilistic prediction results including the predicted mean and predicted variance. The health assessment module is used to run a fuzzy analytic hierarchy process (FAHP) model to assess the current comprehensive health status of the target device and output a comprehensive health assessment score. The HPHI fusion engine module is used to calculate a hybrid predictive health index HPHI based on the outputs of the predictive modeling module and the health assessment module, and by dynamically adjusting the fusion weights using the predictive variance. The dynamic visualization module includes a GPU-accelerated rendering pipeline for rendering the visualization attributes corresponding to HPHI and prediction variance to the target device in the 3D point cloud model in real time.

8. The equipment health system in dynamic operation and maintenance of power distribution networks based on point cloud 3D visualization as described in claim 7. The health assessment system is characterized by: The point cloud management module uses an adaptive octree data structure to organize the 3D point cloud model. The leaf nodes or point cluster nodes of the octree contain pointers to a dynamic data block. The dynamic data block is used to store the time-series data of the device associated with the node, the current HPHI value, and the prediction variance.

9. A computer device, comprising a memory, a processor, and computer-executable instructions stored in the memory, characterized in that: The processor is configured to execute computer-executable instructions to implement the steps of the method according to any one of claims 1-6.

10. A computer-readable storage medium having a program stored thereon, characterized in that: When executed by a processor, the program implements the steps of the method described in any one of claims 1-6.

Images