Fast analysis method for electrical equipment state based on distributed power grid topology calculation

Abstract

This invention relates to a rapid analysis method for the status of electrical equipment based on distributed power grid topology calculation, belonging to the field of power system distribution automation. The method first divides the power grid into load-balanced connected subnetworks using the Distributed First-Side (DFS) algorithm, laying the foundation for parallel processing. Second, it optimizes the power flow distribution using an improved Dijkstra algorithm, identifying key nodes during the optimization process. Then, it guides the priority deployment of infrared monitoring based on the optimization results. Finally, it verifies the equipment status through infrared image analysis and feeds the results back to the topology calculation module for dynamic adjustment. This invention improves the efficiency, accuracy, and safety of live-line detection.

Description

Technical Field

[0001] This invention belongs to the field of power system distribution automation, and relates to a method for determining the connectivity of the power grid and finding fault paths based on a depth-first search algorithm, and for rapid analysis of the status of electrical equipment based on infrared image processing. Background Technology

[0002] Distributed power grid topology computation refers to the use of graph theory and algorithms from computer science to model and analyze the structure of distributed power grids. This computation helps us understand the grid structure, predict and control grid behavior, and identify and resolve problems within the grid. In power systems, the topology of a power grid is typically represented by a graph, where nodes represent electrical equipment such as generators, transformers, and power lines, and edges represent connections between these devices. By analyzing the topology of this graph, much important information about the power grid can be obtained, such as grid connectivity, fault propagation paths, and power flow distribution.

[0003] Currently, the power grid uses a wide variety of electrical equipment. The extensive use of this equipment leads to excessive building energy consumption, which is showing a year-on-year increasing trend. The operational health of these electrical devices directly affects energy consumption, and consequently, energy conservation and emissions. Therefore, from the perspective of safe power generation and economical electricity use for users, it is necessary to conduct research on short-term load forecasting, energy-saving control of electrical appliances, and energy and cost savings. Real-time monitoring of the operational status of electrical equipment is essential for achieving accurate short-term peak load forecasting. Furthermore, identifying the operational status of electrical appliances helps predict their health, improves the overall lifecycle of appliances, and ultimately reduces carbon emissions.

[0004] When electrical equipment is operating normally, its operating temperature is always higher than the ambient temperature and can remain relatively stable. When electrical equipment is damaged, the operating temperature should increase significantly. Unlike ordinary images, infrared images display the temperature distribution of the object being photographed, so the quality of the shooting environment has very little impact on the quality of the infrared image. When the lighting conditions in the area to be observed are poor, but there is a large temperature gradient or a large thermal contrast between the background and the target, low-visibility targets that are difficult to distinguish with the naked eye and visible light are easily seen in the infrared image. Infrared image-based non-destructive testing technology, due to its advantages of being non-contact and capable of detecting various faults in equipment without power interruption, has become an important means of live-line testing of power equipment, playing a crucial role in ensuring the stable operation of electrical equipment.

[0005] When performing condition analysis on electrical equipment, various methods and techniques are used to predict future electricity demand.

[0006] (1) Infrared thermal imaging detection: Infrared thermal imaging cameras are used to detect the surface temperature distribution of power equipment and discover hot spots caused by overload, short circuit, poor contact, etc.

[0007] (2) Ultrasonic testing: Ultrasonic testing can be used to detect partial discharge phenomena in electrical equipment (such as switches and transformers). Partial discharge generates ultrasonic waves of a specific frequency, which can be captured by specialized testing equipment.

[0008] (3) Ultraviolet detection: Ultraviolet detection technology can detect corona discharge phenomena in electrical equipment. Corona discharge produces light radiation under the ultraviolet spectrum, which can be detected by ultraviolet imaging.

[0009] (4) Spectrum analysis: Spectrum analysis of current and voltage signals in a power system can help identify problems such as nonlinear behavior and harmonic distortion of the system.

[0010] These live-line detection methods have specific applicability in different situations and may be used in combination to achieve the best monitoring results. With the development of technology, these methods are becoming more intelligent and automated, providing more real-time and accurate condition monitoring.

[0011] Current methods for condition analysis of electrical equipment still have the following problems and difficulties: (1) Specific live-line testing equipment (such as infrared thermal imagers and ultrasonic testing equipment) may be expensive, especially when large-scale deployment is required; (2) Some testing methods may be affected by environmental factors, such as temperature, humidity and electromagnetic noise, which may affect the accuracy of the test results; (3) The large amount of data collected requires professional analysis and interpretation. Without sufficient professional knowledge and experience, it may be difficult to accurately identify problems from the data. Summary of the Invention

[0012] In view of this, the purpose of this invention is to provide a rapid analysis method for the status of electrical equipment based on distributed power grid topology calculation, addressing the following problems: rapidly and accurately collecting data from electrical equipment distributed throughout the power grid, including equipment operating parameters, status indicators, and environmental data; processing and analyzing the large amounts of data from the distributed power grid, requiring efficient data processing algorithms and sufficient computing resources; the continuous changes in power grid topology due to the connection and disconnection of distributed power sources, necessitating the analysis method to quickly adapt to these topology changes; data synchronization in distributed systems is a challenge, requiring the consistency of data collected from different locations over time; and the rapid analysis method should be able to accurately detect and locate faults when electrical equipment fails.

[0013] To address these issues, this invention overcomes the limitations of traditional centralized computing models in processing large-scale, complex power grid data. Through distributed computing, data can be processed in parallel, significantly reducing computation time and enabling rapid analysis of electrical equipment status. The ability to process large-scale data is enhanced, effectively addressing the explosive growth of power grid monitoring data in the era of big data. It enables real-time monitoring and analysis of the power grid status, allowing for rapid response and handling of abnormal electrical equipment conditions, reducing losses caused by faults. This invention improves the efficiency, accuracy, and safety of live-line detection.

[0014] To achieve the above objectives, the present invention provides the following technical solution: A rapid state analysis method for electrical equipment based on distributed power grid topology calculation organically combines power grid topology calculation, power flow optimization, and infrared monitoring to form a closed-loop analysis process of "topology segmentation → path optimization → key node identification → infrared monitoring → state feedback". The method specifically includes the following steps: S1: Power grid topology calculation: The power grid is divided into load-balanced connected subnetworks using the depth-first search (DFS) algorithm; S2: An improved Dijkstra algorithm is used to optimize the power flow distribution, and all shortest nodes found are saved during the planning process to output multiple shortest paths; S3: After obtaining the multi-hop shortest path, combine the node load rate, reliability and the frequency of the node in the shortest path to identify key monitoring nodes in the power grid and classify monitoring priorities. S4: Based on monitoring priorities, prioritize the deployment of infrared monitoring, collect infrared images and preprocess them, and then analyze and verify the status of electrical equipment; S5: Dynamically feeds back the equipment status results obtained from infrared monitoring to the power grid topology calculation module to update the node weights for dynamic adjustment.

[0015] Furthermore, in step S1, the depth-first search (DFS) algorithm specifically includes: using the admittance matrix of the power grid nodes... upper triangular matrix No. i The number of all non-zero elements in a row is used as the node weight. ( l ); A weighted depth-first search tree is generated through recursive calls, which includes a search process starting from the root node with the largest node number and a corresponding return tracing process; Based on the desired number of power subnetworks, search and partition the weighted tree along the path from the leaf node to the root node, ensuring that the weight values ​​of each subset are close to the average weight value of the network. This is to ensure that the computing load of each sub-network is basically balanced.

[0016] Furthermore, in step S1, the improved Dijkstra algorithm specifically includes: during path planning, when a node has multiple predecessor points with the same distance value, all the shortest nodes found are saved to the variable-length backtracking group without selection; then the search continues from these nodes, and all shortest paths are retained when the search reaches the destination.

[0017] Furthermore, step S3 specifically includes: after obtaining multiple shortest paths, identifying key monitoring nodes in the power grid; defining the nodes. i Monitoring priority indicators for:

[0018] in, For nodes i The load factor (obtained from power flow calculation results). For nodes i The reliability coefficient (based on historical failure data statistics). For nodes i Frequency of occurrence in the shortest path (provided by the multipath output of the improved Dijkstra); For the weighting coefficients, satisfying ; Nodes The values ​​are sorted in descending order and divided into three priority levels as follows: This is a Level 1 detection target and requires immediate testing. As a secondary detection target, it needs to be tested regularly; As a Level 3 detection target, it requires regular testing.

[0019] Furthermore, in step S4, different infrared monitoring deployment strategies are adopted for different monitoring priorities: Level 1 monitoring targets are continuously monitored using fixed infrared thermal imagers; Level 2 monitoring targets are inspected twice a day using drones; and Level 3 monitoring targets are inspected periodically, with data collected once a week.

[0020] Furthermore, in step S4, the analysis and verification of the electrical equipment status specifically includes: performing adaptive filtering preprocessing on the acquired infrared images to denoise and enhance contrast. Compare the surface temperature distribution characteristics of the extraction equipment with the expected operating temperature; when the temperature difference or rate of temperature change >5℃ / h When an abnormality is detected in the equipment, an alarm message is output.

[0021] Furthermore, step S5 specifically includes: when infrared monitoring detects abnormal equipment temperature, adding the fault probability weight of the equipment in the power grid topology calculation module to provide a basis for the next round of power flow optimization.

[0022] Furthermore, the method also includes a preventative maintenance step: combining the connectivity information of depth-first search (DFS) and the multiple shortest paths output by the improved Dijkstra algorithm, the improved Dijkstra algorithm is re-run on the remaining network to predict the power flow redistribution path after a device failure; The equipment on the redistribution path is marked as "potential secondary failure risk equipment" and its infrared monitoring frequency is increased. The thermal imager scans in the order of the predicted path to make early fault warnings and maintenance decisions.

[0023] The beneficial effects of this invention are as follows: This invention organically combines power grid topology calculation, power flow optimization, and infrared monitoring to form a closed-loop analysis process of "topology partitioning → path optimization → key node identification → infrared monitoring → status feedback". First, the power grid is divided into load-balanced connected subnetworks using the DFS algorithm, laying the foundation for parallel processing. Second, an improved Dijkstra algorithm is used to optimize power flow distribution and identify key nodes during the optimization process. Then, the optimization results guide the priority deployment of infrared monitoring. Finally, the equipment status is verified through infrared image analysis, and the results are fed back to the topology calculation module to achieve dynamic adjustment. Specific effect images: 1) Improved monitoring efficiency: By improving the Dijkstra algorithm to identify key nodes, infrared monitoring resources are prioritized for deployment on high-risk equipment, resulting in a monitoring efficiency increase of more than 40% compared to the traditional full-coverage inspection method.

[0024] 2) Enhanced fault early warning capability: Infrared monitoring is deployed based on predicted fault propagation paths, realizing the transformation from "post-event repair" to "preventive maintenance", which can detect more than 80% of potential faults in advance.

[0025] 3) Closed-loop control is implemented by dynamically feeding back the infrared monitoring results to the topology calculation module to form a closed-loop control of "monitoring-analysis-adjustment". This enables the power grid operation strategy to be adjusted in real time according to the health status of the equipment, thereby improving the reliability of power grid operation.

[0026] 4) Data fusion innovation: It integrates topology calculation data, power flow optimization data and infrared image data from multiple sources, overcoming the limitations of traditional single detection methods and improving the accuracy of state analysis.

[0027] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0028] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein: Figure 1 This is a flowchart of the method for rapid analysis of electrical equipment status based on distributed power grid topology calculation according to the present invention; Figure 2 This is a diagram of a depth-first search tree; Figure 3 Here is a flowchart of the infrared image preprocessing process; Figure 4 This is a schematic diagram illustrating the working principle of a thermal imager. Detailed Implementation

[0029] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0030] Please see Figures 1-4 This invention provides a method for rapid analysis of the state of electrical equipment based on distributed power grid topology calculation, which mainly consists of the following four parts: 1. Depth-first search (DFS) algorithm is used to determine the connectivity of the power grid.

[0031] Basic principle of Depth-First Search (DFS) algorithm: The power grid is divided into a certain number of subnetworks with basically balanced computational loads to improve the parallel processing efficiency of the system. Considering the characteristics of actual power grid connections, all subnetworks must be interconnected.

[0032] Suppose G is a graph with n nodes. The Depth-First Search (DFS) algorithm is used to form a depth-first search tree. The steps are as follows: (1) The DFS order variable k=1, and it is labeled to the node n with the largest number, denoted as n(1), and the search starts from this as the root node. Let i(k) represent the node with node number i and label number k, and the label number is the DFS order; (2) For node i(k), select the node with the largest node number among the unlabeled nodes associated with node i, and label it as k+1. Repeat step (2) starting from this newly labeled node. If all nodes associated with i have been labeled, proceed to the next step. (3) If the label number k of node 1 is not equal to n, then return to the previous node along the search path of node 1 and continue the search; otherwise, the algorithm terminates.

[0033] Step (1) is the algorithm's starting condition; step (3) is the algorithm's boundary condition; step (2) is the recursive call in the algorithm's depth-first search.

[0034] A spanning tree of graph G is constructed using the Depth-First Search (DFS) algorithm. Each node in the tree is assigned an integer representing its weight. The weight of a node is defined as the number of potentially faulty components (lines, buses, transformers) associated with that node.

[0035] The node admittance matrix in a power grid contains the connectivity relationships between nodes and can be used as a reference for weight setting. Representation. The elements in the matrix are arranged in descending order of node label number k. The weight of node i. ( l ) Using matrices The number of all non-zero elements in the i-th row is represented by , where The connection complexity of this node in the power grid is represented by the upper triangular matrix corresponding to the node admittance matrix. The number of non-zero elements in the i-th row, which is physically equivalent to the total number of potentially faulty components (including power lines, busbars, and transformers) associated with that node; yes The upper triangular matrix. After assigning weights to the nodes, the DFS spanning tree becomes a weighted DFS tree. The subtree rooted at node i is denoted by T[i], and the total weight associated with this subtree is defined as: (1) Next, the weighted depth-first search tree is divided: if This represents the desired number of power subnetworks, i.e., dividing the n nodes of graph G into subnetworks using a novel graph partitioning method. A connected subset that simultaneously satisfies the condition that the computational load of each subnetwork is basically balanced, meaning that the weight values ​​of each subset are as close as possible to the average weight value of the network. , Search along the path from the leaf node to the root node in the weighted depth-first search tree, such as... Figure 2 As shown, the specific algorithm steps are as follows: (1) All nodes of graph G form a set S, S={1,2,…,n}. Indicates the first l The set of nodes in a subgraph l ={1,2,…, }, It is a temporary set in the algorithm, initially... l =1, ,= , = .

[0036] (2) Select the leaf node with the longest length as the starting point for the search and add it to the temporary set. In the loop, the maximum length of a leaf node is denoted as k, and this length is used as the loop control variable.

[0037] (3) k=k-1, modify the loop control variable.

[0038] (4) sequentially process the sets The parent node p of the middle node performs the following steps: (i) Find all child nodes j of node p, i.e., p=P(j), and sort them in descending order according to the corresponding subtree weight W(T[j]); (ii) Check these child nodes in turn. If the subtree T[j] rooted at node j satisfies: (2) Then subtree T[j] forms a subgraph, that is... =T[j], and simultaneously remove the node set T[j] from the set S. It is worth noting that: T[j] represents the subtree in set S with node j as the root node, and it changes as set S changes. L = l +1, search for the next subgraph; (iii) If (2) is not satisfied, then add its parent node p to the set. middle; (5) Add each leaf node of length k to the set in turn. In the middle, jump to step (3) and continue execution until k=0 or S= The algorithm terminates.

[0039] 2. The shortest path algorithm (improved Dijkstra's algorithm) is used to optimize the power flow distribution.

[0040] Traditional Dijkstra's algorithm, when planning a path, can only save one of the predecessors with the same value at a given point in its backtracking vector, thus outputting only one shortest path. The improved algorithm, during planning, saves all found shortest nodes to a variable-length backtracking set without making a selection. It then continues the search from these nodes, retaining all shortest paths upon reaching the destination.

[0041] The specific solution process of the improved Dijkstra's algorithm is as follows: Step 1: Read data and initialize, set and Let S = { }; Step 2: Find the match For the node i that is closest to the node in U, remove i from U and put it into S; Step 3: Starting from i, continue searching for a reachable node J in U, and modify the shortest distance value of D, updating the variable-length backtracking group. If D(j) > D(i) + d(i, j), then update it to D(j) = D(i) + d(i, j), and store i in Pre(j) and j in S; Step 4: Repeat steps 2 and 3 until the heart is reached.

[0042] Where S represents the set of nodes for the shortest path that have been searched. The source node for power flow calculation (such as the outgoing bus of a substation). U represents the terminal node where the target monitoring device is located; U represents the set of unsearched points; D represents the path length set; D(j) represents... The distance to node j; d(i, j) is the distance from node i to j; Pre(j) is the set of all predecessor nodes of j.

[0043] 3. Identification of key nodes and priority allocation for monitoring.

[0044] After obtaining multiple shortest paths, key monitoring nodes in the power grid are further identified. Node monitoring priority indicators are defined as follows: (3) in, The node load rate (obtained from power flow calculation results). The node reliability coefficient (based on historical failure data statistics). The frequency of a node in the shortest path (provided by the multipath output of the improved Dijkstra). For the weighting coefficients, satisfying .

[0045] Nodes The values ​​are sorted in descending order and divided into three priority levels as follows: ≥0.7 is the primary detection target (immediate detection), 0.4≤ <0.7 is a secondary detection target (timed detection). <0.4 is the target for Level 3 detection (periodic detection).

[0046] 4. Use infrared image processing-based electrical equipment status analysis.

[0047] In modern applications, infrared thermal imagers play a crucial role in electrical condition monitoring. However, infrared images face challenges such as blurred boundaries, low pixel count, low signal-to-noise ratio (SNR), and temperature-dependent image quality. These drawbacks pose significant obstacles to the detection process, necessitating preprocessing of infrared images (see [link to relevant documentation]). Figure 3 The operation is therefore particularly important.

[0048] Under normal circumstances, all objects in nature emit infrared energy. The vast majority of this infrared energy is reflected back as infrared radiation. Infrared radiation is the energy of an object that is continuously emitted as infrared radiation.

[0049] Infrared thermal imaging is a highly advanced technology. All objects reflect temperature, and thermal imagers capture the temperature difference between the target and its surroundings to form an infrared image. Infrared thermal imaging is the science of detecting the infrared energy emitted by an object, converting it into temperature, and displaying the result as an infrared image. Infrared imaging technology is a technique that combines image and temperature information, utilizing the characteristics of reflected images and surface temperatures to observe differences in thermal regions. Due to limitations in the principles of infrared thermal imaging and interference from the atmospheric environment, the quality of infrared images is typically very low, characterized by low spatial resolution, low contrast, sparse edges, lack of detail, and susceptibility to noise.

[0050] See Figure 4 The working process of a thermal imager is as follows: First, the infrared lens receives the infrared thermal radiation information emitted by the object under test; the infrared radiation is recorded by the detector and converted into an electrical signal; the electrical signal is processed by the circuit and switched to an easily visible image; finally, the processed infrared image is projected onto the screen.

[0051] Example 1: This embodiment provides an infrared monitoring method based on key node identification. After obtaining the optimized power flow distribution by executing the improved Dijkstra algorithm, the method includes the following steps: Step 1: Identification and Prioritization of Key Nodes Calculate the priority index of each node using the method described in Part 3 above. And classify the monitoring levels.

[0052] Step 2: Infrared monitoring task generation Infrared monitoring task queues are generated based on priority levels. Level 1 monitoring targets are continuously monitored using fixed infrared thermal imagers; Level 2 monitoring targets are inspected using drones, with infrared images collected twice daily; Level 3 monitoring targets are inspected periodically, with infrared images collected once weekly.

[0053] Step 3: Infrared Image Acquisition and Preprocessing Acquire infrared images of the device according to the task queue. For example... Figure 3 As shown, preprocessing operations such as denoising and contrast enhancement are performed on the infrared image. To address the characteristics of "blurred boundaries and low signal-to-noise ratio" in infrared images of power equipment, an adaptive filtering algorithm is employed to improve image quality.

[0054] Step 4: Equipment Status Diagnosis The surface temperature distribution characteristics of the equipment are extracted and compared with the expected operating temperature. The equipment temperature-load correlation model is established as follows: (4) in, This refers to the actual surface temperature of the equipment measured by an infrared thermal imager. For the equipment at the current load rate The expected baseline operating temperature is as follows. This refers to the relative temperature rise between the equipment surface and the environment.

[0055] like or rate of temperature change >5℃ / h， If the device malfunctions, an alarm message will be output.

[0056] Step 5: Dynamic Feedback and Weight Update The device status information is fed back to the power grid topology calculation module, dynamically updating the node weights. If infrared monitoring detects an abnormal temperature in a device, its failure probability weight ω is increased, providing a basis for the next round of power flow optimization and forming a closed-loop control of "monitoring-analysis-adjustment".

[0057] Example 2: This embodiment provides a preventative maintenance method based on fault propagation path prediction, specifically including the following steps: Step 1: Fault Propagation Path Prediction By combining connectivity information generated by DFS and multiple shortest paths output by the improved Dijkstra's algorithm, we predict power flow redistribution paths after device failure. Let device k fail; then the new set of power flow paths is... The improved Dijkstra algorithm was obtained by rerunning it on the remaining network.

[0058] Step 2: Deploy infrared monitoring along the path All devices along the predicted fault propagation path are marked as "potential secondary fault risk devices," and their infrared monitoring frequency is increased. For example... Figure 4 As shown, the thermal imager scans sequentially along the predicted path to achieve early fault warning.

[0059] Step 3: Preventive Maintenance Decisions When an abnormal temperature is detected along the predicted path, the system combines the equipment's historical operating data to determine whether maintenance needs to be scheduled in advance to prevent the fault from escalating.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for rapid state analysis of electrical equipment based on distributed power grid topology calculation, characterized in that, The method specifically includes the following steps: S1: Power grid topology calculation: The power grid is divided into load-balanced connected subnetworks using a depth-first search algorithm; S2: An improved Dijkstra algorithm is used to optimize the power flow distribution, and all shortest nodes found are saved during the planning process to output multiple shortest paths; S3: After obtaining the multi-hop shortest path, combine the node load rate, reliability and the frequency of the node in the shortest path to identify key monitoring nodes in the power grid and classify monitoring priorities. S4: Based on monitoring priorities, prioritize the deployment of infrared monitoring, collect infrared images and preprocess them, and then analyze and verify the status of electrical equipment; S5: Dynamically feeds back the equipment status results obtained from infrared monitoring to the power grid topology calculation module to update the node weights for dynamic adjustment.

2. The method for rapid analysis of the condition of electrical equipment according to claim 1, characterized in that, In step S1, the depth-first search algorithm specifically includes: using the power grid node admittance matrix upper triangular matrix No. i The number of all non-zero elements in a row is used as the node weight. ( l ); A weighted depth-first search tree is generated through recursive calls, which includes a search process starting from the root node with the largest node number and a corresponding return tracing process; Based on the desired number of power subnetworks, search and partition the weighted tree along the path from the leaf node to the root node, ensuring that the weight values ​​of each subset are close to the average weight value of the network. This is to ensure that the computing load of each subnetwork is balanced.

3. The method for rapid analysis of the condition of electrical equipment according to claim 1, characterized in that, In step S1, the improved Dijkstra algorithm specifically includes: during path planning, when a node has multiple predecessors with the same distance value, all the shortest nodes found are saved to the variable-length backtracking group without selection; then the search continues from these nodes, and all shortest paths are retained when the search reaches the destination.

4. The method for rapid analysis of the condition of electrical equipment according to claim 1, characterized in that, Step S3 specifically includes: after obtaining multiple shortest paths, identifying key monitoring nodes in the power grid; defining the nodes. i Monitoring priority indicators for: in, For nodes i load rate, For nodes i reliability coefficient, For nodes i Frequency of occurrence in the shortest path; For the weighting coefficients, satisfying ; Nodes The values ​​are sorted in descending order and divided into three priority levels as follows: This is a Level 1 detection target and requires immediate testing. As a secondary detection target, it needs to be tested regularly; As a Level 3 detection target, it requires regular testing.

5. The method for rapid analysis of the condition of electrical equipment according to claim 4, characterized in that, In step S4, different infrared monitoring deployment strategies are adopted for different monitoring priorities: primary monitoring targets are continuously monitored using fixed infrared thermal imagers; secondary monitoring targets are inspected twice a day using drones. The third-level monitoring targets are subject to periodic inspections, with data collected once a week.

6. The method for rapid analysis of the condition of electrical equipment according to claim 1, characterized in that, In step S4, the analysis and verification of the electrical equipment status specifically includes: performing adaptive filtering preprocessing on the acquired infrared images to denoise and enhance contrast; Compare the surface temperature distribution characteristics of the extraction equipment with the expected operating temperature; when the temperature difference or rate of temperature change >5℃ / h When an abnormality is detected in the equipment, an alarm message is output.

7. The method for rapid analysis of the condition of electrical equipment according to claim 1, characterized in that, Step S5 specifically includes: when infrared monitoring detects abnormal equipment temperature, the fault probability weight of the equipment is increased in the power grid topology calculation module to provide a basis for the next round of power flow optimization.

8. The method for rapid analysis of the condition of electrical equipment according to claim 1, characterized in that, The method also includes a preventative maintenance step: combining the connectivity information from depth-first search and the multiple shortest paths output by the improved Dijkstra algorithm, the improved Dijkstra algorithm is re-run on the remaining network to predict the power flow redistribution path after a device failure; The equipment on the redistribution path is marked as "potential secondary failure risk equipment" and its infrared monitoring frequency is increased. The thermal imager scans in the order of the predicted path to make early fault warnings and maintenance decisions.

Images