Power grid operation state prediction method, device, equipment and medium

Abstract

The invention discloses a power grid operation state prediction method and device, equipment and a medium, and the method comprises the following steps: building a meteorological grid with an equipment node as a center, and obtaining tree data in the meteorological grid and meteorological data in the meteorological grid; based on the tree data and the suspension data of the connection lines of the equipment nodes in the meteorological grid, determining a toppling risk probability through a toppling tree database; matching the toppling risk probability and the real-time meteorological data with a preset rule base to determine risk factors; and completing prediction of the power grid operation state according to the risk factor and a preset mapping relation. By establishing a meteorological grid and acquiring tree data, comprehensive analysis and evaluation of an external environment are realized by adopting a mode of combining a tree threat value and a toppled tree database, changes and risk evolution of a power grid operation condition can be reflected in time, and an overall operation risk faced by a power grid can be accurately identified and evaluated; and the accuracy of predicting the operation state is improved.

Description

Technical Field

[0001] This invention relates to the field of power grid operation prediction technology, and in particular to a method, apparatus, equipment and medium for predicting power grid operation status. Background Technology

[0002] Currently, the power grid, as the lifeline of energy transmission in modern society, plays an important role in social development. Among them, the prediction of the power grid's operating status is crucial for its stable operation. However, in mountainous and rural areas where power grids are erected, they are frequently subject to various faults due to wind disasters, such as trees falling and knocking down power poles, or old power poles being snapped by wind.

[0003] Therefore, Chinese invention patent CN114564889B discloses a method for early warning of power distribution network wind disasters based on the PCA model. This method obtains ontological parameters, refined meteorological parameters, and operational status parameters at the time of a fault from a historical fault sample database to form a training sample set. It then uses a multivariate regression method to fit the outage duration of the training sample set, establishing a power outage duration prediction model based on multivariate regression. Using the PCA model, it obtains the principal components of relevant multivariate parameters during the multivariate regression training process and extracts the principal component parameters to establish a PCA-based power outage duration prediction model for power distribution network wind disasters. Finally, it performs SVM classification on the power outage duration of power distribution network wind disasters. Based on the trained prediction model and SVM classification, it provides early warning of power distribution network wind disasters.

[0004] However, the above-mentioned technologies only analyze the duration of power outages caused by faults and the physical parameters, refined meteorological parameters, and operational status parameters at the time of the fault. Since the fault-inducing factors such as terrain and trees at different distribution network equipment nodes are different, they lack a comprehensive analysis and evaluation of the external environment on the power grid, and the accuracy of data early warning is not high. Summary of the Invention

[0005] To address the technical problems of existing power grid risk early warning systems, such as lack of comprehensive analysis of the external environment, insufficient parameter settings in prediction methods, incomplete risk assessment methods, lack of threshold adjustment mechanisms, and insufficient real-time monitoring and dynamic adjustment mechanisms, and to achieve accurate prediction and risk assessment of power grid operating status, this invention provides a power grid operating status prediction method, device, equipment, and medium.

[0006] The technical solution adopted by the present invention to solve its technical problem provides a method for predicting the operating status of a power grid, including: establishing a meteorological grid centered on equipment nodes, and acquiring tree data and meteorological data within the meteorological grid; Based on the tree data and the suspension data of the connection lines of the equipment nodes in the meteorological grid, the probability of falling risk is determined by the falling tree database according to the meteorological data; The probability of dumping risk and the meteorological data are matched with a preset rule base to determine the risk factors; The prediction of the power grid's operating status is completed based on the risk factors and preset mapping relationships.

[0007] Optionally, the step of establishing a meteorological grid centered on device nodes and acquiring tree data and meteorological data within the meteorological grid includes: Obtain the geographic coordinates of each device node within the target area; The meteorological data within the target area is divided into grids to obtain a meteorological grid; The meteorological grid where the device node is located is determined based on the geographic coordinates, and the meteorological grid where the device node is located is used as the target meteorological grid. The meteorological data corresponding to the target meteorological grid is recorded. Identify trees within the target meteorological grid that are located within a preset range centered on the connection line of the device node, and determine the tree data of the trees, which includes tree species, tree diameter at breast height (DBH), tree crown shape, tree location, and tree height.

[0008] Optionally, the step of determining the probability of falling trees based on the tree data and the suspension data of the connection lines of the equipment nodes within the meteorological grid, using the meteorological data through the fallen tree database, includes: Based on the tree location and tree height, and in conjunction with the suspension data of the connecting lines, the tree threat value is determined, and the tree hazard factor is determined based on the tree threat value; Based on the tree species category, the tree diameter at breast height (DBH) and the tree crown shape data, the probability of tree falling is determined by the fallen tree database according to the meteorological data, and the product of the tree hazard factor and the tree falling probability is recorded as the falling risk probability.

[0009] Optionally, the step of determining a tree threat value based on the tree location and tree height, combined with the suspension data of the connecting line, and determining a tree hazard factor based on the tree threat value, includes: Based on the tree location and the suspension data, the horizontal distance between the tree and the connecting line is calculated, and the difference between the horizontal distance and the tree height is calculated to obtain the horizontal threat value; Determine whether the level of threat is below a first threshold; If the horizontal threat value is lower than the first threshold, the difference between the line suspension height and the tree height is calculated based on the tree height and the suspension data to obtain the height threat value; Determine whether the high threat value is lower than the second threshold; If the height threat value is lower than the second threshold, the sum of the height threat value and the horizontal threat value is calculated to obtain the total threat value, and the ratio of the height threat value to the total threat value is used as the tree risk factor.

[0010] Optionally, the step of determining the probability of tree falling based on the tree species category, the tree diameter at breast height (DBH), and the tree crown shape data, using the meteorological data and the fallen tree database, includes: Cluster analysis was performed on the fallen tree data in the fallen tree database according to the wind level data in the meteorological data to obtain multiple sets of fallen tree characteristics; The tree species category, tree diameter at breast height (DBH), and tree crown shape data are compared with multiple sets of features of fallen trees using Gower similarity calculation to generate a first similarity score. The first similarity scores are then sorted in descending order, and the first similarity score with the largest value is selected as the probability of tree falling.

[0011] Optionally, the step of identifying trees within the target meteorological grid that are located within a preset range centered on the connection line of the device node, and determining the tree data of the trees, includes: Obtain point cloud data of trees within a preset range with the connection line of the device node as the center line within the target meteorological grid, and determine the incomplete data of the trees; Using tree species, diameter at breast height (DBH), height, and crown shape as features, complete data that is closest to the incomplete data is selected from the historical database and used as template data. The template data and the incomplete data are horizontally sliced ​​at the same height to obtain data slices, and the perimeter points of each data slice are counted to obtain point count statistics. Based on the point count statistics, the data slices are sorted in ascending order by the count difference. The data slice with the smallest difference is selected as the baseline layer. The template data is scaled proportionally in the horizontal and vertical directions according to the baseline layer so that the perimeter points of the template data and the incomplete data are completely consistent at the baseline layer. At the edge of the missing region of the incomplete data, take three adjacent points in a clockwise direction to form a triangle. Calculate the displacement vector from the center of the triangle to the corresponding position of the template data. Apply the displacement vector to the corresponding point of the template data to make the corresponding region of the triangle translate to the missing region of the incomplete data until the missing region is completely covered by the continuous surface formed by the triangle, and obtain the completed incomplete data. The incomplete data is sliced ​​using contour lines, and the number of perimeter points in each slice is counted. The percentage difference is calculated based on the number of perimeter points in the incomplete data slices without missing points. If the percentage difference exceeds a preset threshold, the process continues with the step of acquiring point cloud data of trees within a preset range centered on the connection line of the device node within the target meteorological grid, and determining the incomplete data of the trees.

[0012] Optionally, the step of matching the dumping risk probability and the meteorological data with a preset rule base to determine the hazard factors includes: The historical number of faults is counted in the historical database based on the meteorological data, and the selected value is determined based on the historical number of faults. The probability of tipping over and the meteorological data are matched with a preset rule base to determine the initial hazard factor. The judgment rules in the rule base include: if the wind speed is greater than 8 m / s and the probability of tipping over is greater than 60, the risk factor increases by 0.3; if the humidity is greater than 80% and the temperature is greater than 30℃, the risk factor increases by 0.2. Based on the selected value, the initial hazard factor is corrected to obtain the first hazard factor among the hazard factors; The device node whose first hazard factor is greater than or equal to the third threshold is taken as the target node, and the device node directly adjacent to the target node is taken as the neighbor node. The number of cascading failures and the total number of failures are determined based on historical databases, and the ratio of the number of cascading failures to the total number of failures is used as the diffusion coefficient. The risk factors of adjacent nodes are calculated based on the diffusion coefficient, and the risk factors of adjacent nodes are used as the second risk factor among the risk factors.

[0013] On the other hand, this application provides a power grid operation status prediction device, the device comprising: The data acquisition module is used to establish a meteorological grid centered on the device nodes and to acquire tree data and meteorological data within the meteorological grid. The probability prediction module is used to determine the probability of falling trees based on the tree data and the suspension data of the connection lines of the equipment nodes in the meteorological grid, and to determine the probability of falling trees through the falling tree database according to the meteorological data. The matching module is used to match the dumping risk probability and the meteorological data with a preset rule base to determine the hazard factors; The prediction module is used to predict the operating status of the power grid based on hazard factors and preset mapping relationships.

[0014] Thirdly, this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the power grid operation state prediction method as described above.

[0015] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the power grid operation state prediction method described above.

[0016] The beneficial effects of this invention are: This application provides a power grid operation status prediction method. By establishing a meteorological grid and acquiring tree data, it achieves a comprehensive analysis and assessment of the external environment, overcoming the limitations of traditional power grid risk early warning methods that rely solely on sensors and monitoring systems. This improves the accuracy of predicting operation status. By combining tree threat values ​​and a database of fallen trees, it comprehensively considers external environmental factors such as tree obstacles, accurately identifying and assessing the overall operational risks faced by the power grid, thus improving the reliability of power grid operation safety. Through matching real-time meteorological data with a rule base and combining analysis of historical databases of equipment nodes, it achieves real-time monitoring and dynamic adjustment of the power grid operation status, enabling timely reflection of changes in power grid operation status and risk evolution. By employing an adaptive rule base and threshold mechanism, it dynamically adjusts risk factors according to real-time meteorological conditions, overcoming the limitations of fixed thresholds in existing technologies and improving the accuracy and reliability of association rules. Through traversal of the adjacency matrix and diffusion of risk factors, it achieves intelligent propagation and risk sharing of power grid equipment operation status, enhancing the overall safety and stability of power grid operation. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the operation of a power grid operation status prediction method according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a power grid operation status prediction device according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of this application. Detailed Implementation

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0019] A method for predicting the operating status of a power grid, comprising: S101. Establish a meteorological grid centered on the device nodes, and acquire tree data and meteorological data within the meteorological grid; In one possible implementation, the step of establishing a meteorological grid centered on device nodes and acquiring tree data and meteorological data within the meteorological grid includes: Obtain the geographic coordinates of each device node within the target area; The meteorological data within the target area is divided into grids to obtain a meteorological grid; The meteorological grid where the device node is located is determined based on the geographic coordinates, and the meteorological grid where the device node is located is used as the target meteorological grid. The meteorological data corresponding to the target meteorological grid is recorded. Identify trees within the target meteorological grid that are located within a preset range centered on the connection line of the device node, and determine the tree data of the trees, which includes tree species, tree diameter at breast height (DBH), tree crown shape, tree location, and tree height.

[0020] For example, the geographical coordinates of each device node are obtained, and an adjacency matrix is ​​generated based on the connection relationships. Device nodes in the power grid include substations, distribution cabinets, power poles, etc. By obtaining the latitude and longitude coordinates of these devices, connection relationships between device nodes are established, forming an adjacency matrix. The adjacency matrix represents the connection status between each device node; if there is a direct connection between two nodes, the corresponding position in the matrix has a value of 1, otherwise it is 0. Meteorological data is gridded, and device nodes are bound to corresponding meteorological grids based on their geographical coordinates. The entire power grid coverage area is divided into multiple meteorological grids, each with a size that can be set to 1km × 1km or adjusted according to actual needs. This ensures localized refinement of meteorological data, adapting to different regional terrain and microclimate characteristics. Based on the geographical coordinates of each device node, it is mapped to the corresponding meteorological grid, achieving a one-to-one correspondence between device nodes and meteorological grids, laying the foundation for subsequent accurate association of meteorological risks with device nodes. The locations and corresponding heights of trees within the warning range on both sides of the connecting lines within the meteorological grid are obtained, enabling precise location of tipping risks. By using technologies such as lidar scanning or satellite remote sensing, information on trees within the warning range on both sides of the power line in each meteorological grid (usually a range of 10 meters on both sides of the line) is obtained, including the geographical coordinates of the trees and tree height data, providing three-dimensional spatial data with millimeter-level accuracy, which significantly improves the accuracy of tree location and height measurement.

[0021] S102. Based on the tree data and the suspension data of the connection lines of the equipment nodes in the meteorological grid, determine the probability of falling risk according to the meteorological data through the fallen tree database; In one possible implementation, the step of determining the probability of falling trees based on the tree data and the suspension data of the connection lines of the equipment nodes within the meteorological grid, using the meteorological data through a fallen tree database, includes: Based on the tree location and tree height, and in conjunction with the suspension data of the connecting lines, the tree threat value is determined, and the tree hazard factor is determined based on the tree threat value; Based on the tree species category, the tree diameter at breast height (DBH) and the tree crown shape data, the probability of tree falling is determined by the fallen tree database according to the meteorological data, and the product of the tree hazard factor and the tree falling probability is recorded as the falling risk probability.

[0022] For example, the location and height of each tree within the warning range on both sides of the connecting line in the meteorological grid are obtained, and the tree threat value is determined by combining the hanging height of the connecting line; the probability of tree falling is determined based on the database of fallen trees, and the product of the tree hazard factor and the probability of tree falling is recorded as the probability of falling risk; thereby, this application integrates the physical distance threat and the meteorological disaster probability to generate a more comprehensive single-point falling risk assessment.

[0023] In one possible implementation, the step of determining a tree threat value based on the tree location and tree height, combined with suspension data of the connecting line, and determining a tree hazard factor based on the tree threat value, includes: Based on the tree location and the suspension data, the horizontal distance between the tree and the connecting line is calculated, and the difference between the horizontal distance and the tree height is calculated to obtain the horizontal threat value; Determine whether the level of threat is below a first threshold; If the horizontal threat value is lower than the first threshold, the difference between the line suspension height and the tree height is calculated based on the tree height and the suspension data to obtain the height threat value; Determine whether the high threat value is lower than the second threshold; If the height threat value is lower than the second threshold, the sum of the height threat value and the horizontal threat value is calculated to obtain the total threat value, and the ratio of the height threat value to the total threat value is used as the tree risk factor.

[0024] For example, the location, height, and suspension height of trees within the warning range on both sides of the connecting line are extracted. For each tree, its geographical coordinates, height, and the corresponding suspension height of the line are recorded.

[0025] Calculate the horizontal distance between the tree's location and the connecting road, and then calculate the difference between this horizontal distance and the tree's height to obtain a horizontal threat value. Determine if the fall threat value is below a first threshold. The horizontal threat value represents the horizontal distance margin by which a tree might fall and touch the road. A smaller value indicates a higher likelihood of the tree touching the road. The first threshold can be set to 5 meters. If the horizontal threat value is below 5 meters, proceed to the next step.

[0026] If the horizontal threat value is below the first threshold, the difference between the line hanging height and the tree height is calculated to obtain the height threat value, and it is then determined whether it is below the second threshold. The height threat value represents the difference in vertical distance between the tree and the line. When this value is small or negative, it indicates that the tree height is close to or exceeds the line height. The second threshold can be set to 3 meters. When the height threat value is below 3 meters, the next calculation is performed.

[0027] If the height threat value is below the second threshold, the tree hazard factor is calculated as the ratio of the height threat value to the sum of its own height threat value and the horizontal threat value. The formula for calculating the tree hazard factor is: Tree Hazard Factor = Height Threat Value / (Height Threat Value + Horizontal Threat Value). The closer this value is to 1, the greater the threat the tree poses to the line.

[0028] In one possible implementation, the step of determining the probability of tree falling based on the tree species category, the tree diameter at breast height (DBH), and the tree crown shape data, according to the meteorological data and the fallen tree database, includes: Cluster analysis was performed on the fallen tree data in the fallen tree database according to the wind level data in the meteorological data to obtain multiple sets of fallen tree characteristics; The tree species category, tree diameter at breast height (DBH), and tree crown shape data are compared with multiple sets of features of fallen trees using Gower similarity calculations to generate a first similarity score. These first similarity scores are then sorted in descending order, and the score with the highest value is recorded as the tree's probability of falling. Gower similarity is a similarity metric used to process heterogeneous datasets. It can integrate various types of variables, including continuous and discrete data, and is widely used in cluster analysis, reliability screening, and other fields.

[0029] For example, a fallen tree database is accessed, and using the wind level from real-time meteorological data as the standard, key characteristics of trees causing damage under different wind levels are mined based on historical big data. Cluster analysis is then performed on the fallen tree data in the database to obtain multiple sets of fallen tree characteristics. The fallen tree database records the historical tree falling under different wind conditions, including tree species, diameter at breast height (DBH), and crown shape data. Through cluster analysis, characteristic groups of trees prone to falling under different wind levels can be obtained.

[0030] Tree point cloud data from lidar scans and historical inspection records were extracted to obtain tree data for each tree within the warning range on both sides of the connecting line. The tree data includes tree species, diameter at breast height (DBH), and crown shape data, which were obtained through lidar scans and historical inspection records.

[0031] The Gower similarity test is performed sequentially on the tree data of each tree and the characteristics of multiple sets of fallen trees to generate a first similarity score. These first similarity scores are then sorted in descending order, and the score with the highest value is recorded as the probability of the tree falling. This serves to quantitatively assess the likelihood of a single tree falling under a specific wind level. Gower similarity is a similarity calculation method that can be used for mixed data types, simultaneously handling tree species (categorical data), tree diameter at breast height (DBH) (continuous data), and tree crown shape (multidimensional data). A higher similarity score indicates a greater similarity between the tree and historically fallen trees, and thus a higher probability of falling.

[0032] S103. Match the dumping risk probability and the meteorological data with a preset rule base to determine the risk factors; In one possible implementation, the step of matching the dumping risk probability and the meteorological data with a preset rule base to determine the hazard factor includes: The historical number of faults is counted in the historical database based on the meteorological data, and the selected value is determined based on the historical number of faults. The probability of tipping over and the meteorological data are matched with a preset rule base to determine the initial hazard factor. The judgment rules in the rule base include: if the wind speed is greater than 8 m / s and the probability of tipping over is greater than 60, the risk factor increases by 0.3; if the humidity is greater than 80% and the temperature is greater than 30℃, the risk factor increases by 0.2. Based on the selected value, the initial hazard factor is corrected to obtain the first hazard factor among the hazard factors; The device node whose first hazard factor is greater than or equal to the third threshold is taken as the target node, and the device node directly adjacent to the target node is taken as the neighbor node. The number of cascading failures and the total number of failures are determined based on historical databases, and the ratio of the number of cascading failures to the total number of failures is used as the diffusion coefficient. The risk factors of adjacent nodes are calculated based on the diffusion coefficient, and the risk factors of adjacent nodes are used as the second risk factor among the risk factors.

[0033] For example, real-time meteorological data of the device nodes is obtained, including temperature, humidity, and wind speed. Real-time meteorological information of the location of each device node is obtained through a meteorological station or meteorological data service, including the current temperature (°C), humidity (%), and wind speed (m / s).

[0034] Obtain all tipping risk probabilities within the meteorological grid corresponding to each device node, and select the maximum value. For each device node, find all calculated tipping risk probabilities within its corresponding meteorological grid, and select the maximum value as the representative tipping risk probability value for that node.

[0035] Retrieve the historical failure counts of the device node under the same meteorological conditions from the historical database. Query the historical database to count the historical failure counts of this device node under similar meteorological conditions (similar temperature, humidity, and wind speed ranges).

[0036] Configure a rule base and execute the judgment rules within it. The rule base updates data and executes rule judgments at preset intervals, for example, updating data and re-executing the judgment every 30 minutes. The judgment rules within the rule base include: If the wind speed is greater than 8 m / s and the probability of tipping over is greater than 60, the risk factor increases by 0.3. If the humidity is greater than 80% and the temperature is greater than 30℃, the risk factor increases by 0.2; The risk factor is adjusted, and the adjustment value is selected as one-tenth of the historical number of failures.

[0037] The upper limit of the risk factor is set to 1.0. The incrementally accumulated risk factor is set as the base value of the risk factor, and the risk factor of the corresponding device node in the adjacency matrix is ​​updated. For each device node, the risk factor is calculated based on the rule base judgment result, ensuring that the risk factor does not exceed 1.0. The calculated risk factor is then updated to the position of the corresponding device node in the adjacency matrix.

[0038] The geographical coordinates of each device node are obtained, and an adjacency matrix is ​​generated based on the connection relationships of each device node. The adjacency matrix is ​​traversed, and for device nodes whose risk factors exceed a first threshold, the risk factors are propagated to their neighboring device nodes. The risk factors of the corresponding device nodes in the adjacency matrix are then updated a second time. The first threshold can be set to 0.7. When the risk factor of a device node exceeds 0.7, the risk factor will be propagated to its neighboring nodes. In this way, this application simulates the cascading propagation effect of power grid faults and improves the ability to predict systemic risks.

[0039] For device nodes whose initial risk factor is greater than the first threshold, find and record their directly adjacent device nodes. The adjacency matrix can then be used to quickly locate other device nodes directly connected to the high-risk node.

[0040] Historical databases are obtained to record the number of cascading failures and the total number of failures. The ratio of cascading failures to total failures is used as the diffusion coefficient. A cascading failure refers to a situation where a failure in one device leads to a failure in adjacent devices. The probability of cascading failures can be obtained through historical data analysis, which serves as the risk diffusion coefficient. Therefore, this application quantifies the intensity of risk propagation based on historical data of actual power grid operation, thereby improving the accuracy of the diffusion model.

[0041] The risk factor of adjacent equipment nodes increases by multiplying itself by the diffusion coefficient; the upper limit of the risk factor is limited to 1.0. For example, if a node has a risk factor of 0.8 and a diffusion coefficient of 0.5, the risk factor of its adjacent nodes will increase by 0.8 × 0.5 = 0.4, but the total risk factor will not exceed 1.0 to prevent the risk value from being infinitely amplified and to ensure the reasonableness of the results.

[0042] S104. Based on the risk factors and preset mapping relationships, complete the prediction of the power grid operation status.

[0043] For example, a preset mapping relationship between risk factors and operating states is established to predict operating states. The value range of risk factors is divided, with risk factors below a first threshold matched with the normal operating state of the power grid; risk factors above the first threshold are matched with the warning operating state of the power grid. For instance, when the risk factor is between 0 and 0.7, it corresponds to the normal operating state of the power grid; when the risk factor is between 0.7 and 1.0, it corresponds to the warning operating state of the power grid. Operating state prediction is triggered based on the risk factors. The system predicts the operating state of each part of the power grid based on the final risk factor value of each device node and the preset mapping relationship, and issues corresponding level warning information to maintenance personnel, indicating potential fault risks.

[0044] By comprehensively considering the risk of tipping over, meteorological conditions, and historical fault data, the power grid operating status can be accurately predicted, providing an effective early warning mechanism for the safe operation of the power grid.

[0045] By establishing a meteorological grid and acquiring tree data, a comprehensive analysis and assessment of the external environment is achieved, overcoming the limitations of traditional power grid risk early warning systems that rely solely on sensors and monitoring systems. This improves the accuracy of predicting operational status. By combining tree threat values ​​and a database of fallen trees, external environmental factors such as tree obstacles are comprehensively considered, accurately identifying and assessing the overall operational risks faced by the power grid, thus improving the reliability of power grid operation safety. Through matching real-time meteorological data with a rule base and analyzing historical databases of equipment nodes, real-time monitoring and dynamic adjustment of the power grid's operational status are achieved, enabling timely reflection of changes in the power grid's operational status and risk evolution. Adaptive rule bases and threshold mechanisms are used to dynamically adjust risk factors based on real-time meteorological conditions, overcoming the limitations of fixed thresholds in existing technologies and improving the accuracy and reliability of association rules. Through traversal of the adjacency matrix and diffusion of hazard factors, intelligent propagation and risk sharing of power grid equipment operational status are achieved, enhancing the overall safety and stability of power grid operation.

[0046] In one possible implementation, the step of identifying trees within a preset range centered on the connection line of the device node within the target meteorological grid, and determining the tree data of the trees, includes: Obtain point cloud data of trees within a preset range with the connection line of the device node as the center line within the target meteorological grid, and determine the incomplete data of the trees; Using tree species, diameter at breast height (DBH), height, and crown shape as features, complete data that is closest to the incomplete data is selected from the historical database and used as template data. The template data and the incomplete data are horizontally sliced ​​at the same height to obtain data slices, and the perimeter points of each data slice are counted to obtain point count statistics. Based on the point count statistics, the data slices are sorted in ascending order by the count difference. The data slice with the smallest difference is selected as the baseline layer. The template data is scaled proportionally in the horizontal and vertical directions according to the baseline layer so that the perimeter points of the template data and the incomplete data are completely consistent at the baseline layer. At the edge of the missing region of the incomplete data, take three adjacent points in a clockwise direction to form a triangle. Calculate the displacement vector from the center of the triangle to the corresponding position of the template data. Apply the displacement vector to the corresponding point of the template data to make the corresponding region of the triangle translate to the missing region of the incomplete data until the missing region is completely covered by the continuous surface formed by the triangle, and obtain the completed incomplete data. The incomplete data is sliced ​​using contour lines, and the number of perimeter points in each slice is counted. The percentage difference is calculated based on the number of perimeter points in the incomplete data slices without missing points. If the percentage difference exceeds a preset threshold, the process continues with the step of acquiring point cloud data of trees within a preset range centered on the connection line of the device node within the target meteorological grid, and determining the incomplete data of the trees.

[0047] For example, since the trees on both sides of the road appear in patches, when scanning them individually, whether it is an aerial 3D scan by a drone or a 3D scan by a ground vehicle, data will inevitably be incomplete, which will result in the inability to obtain accurate data on the trees.

[0048] Step a: Extract the tree point cloud data from the lidar scan and obtain the incomplete data of each tree within the warning range on both sides of the connecting line; if the diameter at breast height is missing, use "trunk ellipticity" as the fifth sorting key to ensure that the template and the target cross-section shape are similar. Step b: Call the historical database and select the complete data that is closest to the incomplete data in the historical database; the closest complete data is obtained by using tree species, tree diameter at breast height, tree height and tree crown shape as features, and calculate the Gower similarity between the complete data and the incomplete 3D data respectively, and select the complete data corresponding to the maximum value as the template data; Step c: Quantitatively slice the template data and the incomplete data horizontally with equal height. Count the perimeter points of each slice and sort them in ascending order based on the difference in counts. Select the slice with the smallest difference as the baseline layer. Using the baseline layer as a reference, scale the overall outline of the template data proportionally in both the horizontal and vertical directions to ensure that the template data and the incomplete data have the same perimeter point count at the baseline layer. Perform "nearest point pairing" on the scaled template data: calculate the minimum distance to the existing edge points of the incomplete data. If the distance is less than a preset threshold, retain the data; otherwise, discard it, achieving initial alignment. When counting the perimeter points, the tree is horizontally cut at a certain height to obtain the edge points of a cross section; the total number of points counted along this edge is the "perimeter points" of that layer. When calculating the difference, first obtain the "perimeter points" of the template tree at the corresponding height, then obtain the "perimeter points" of the target tree at the same height; subtract the two, comparing only "how many more or fewer points," and the absolute value of this "how many more or fewer points" is the "counting difference." For example, if the template data has 100 edge points at a certain layer, and the incomplete data has 95 edge points at the same layer, then the counting difference is 5. Step d: At the edge of the missing area of ​​the incomplete data, take three adjacent points in a clockwise direction to form a triangle, calculate the displacement vector from the center of the triangle to the corresponding position of the template data, and apply the displacement vector to the corresponding point of the template data so that the triangle as a whole is translated to the missing area of ​​the incomplete data. That is, first find the current position of the small curved surface (the small triangle formed by the three points just now) on the "template data"; then, like moving building blocks, move this small triangle along with its surrounding neighboring points to the missing part on the "incomplete data" so that the center point of the triangle is exactly touching the center point of the missing edge. In this process, applying the displacement vector to the corresponding point of the template data means that all the points to be moved from the template tree follow the same route as this drawing—each point moves three steps to the left and two steps forward. In other words, the small curved surface of the entire template data is moved to the gap position of the incomplete data to complete the filling. Step e: Repeat step d above until the missing area is completely covered by a continuous surface formed by multiple triangles; Step f: If holes appear, repeat steps d and e on the edges of the holes until there are no remaining holes in the residual data or the maximum number of iterations is reached. Step g: The completed incomplete data is then sliced ​​again to the same height. The number of perimeter points in each slice is counted. The percentage difference is calculated based on the number of perimeter points in the original incomplete data slice without missing points. If the percentage difference in any layer exceeds the preset threshold, return to step a, reselect the suboptimal template and repeat the subsequent steps. When the percentage difference in all slices is lower than the preset threshold, the final completion result is output.

[0049] On the other hand, such as Figure 2 As shown, this application provides a power grid operation status prediction device, the device comprising: The data acquisition module 201 is used to establish a meteorological grid centered on the device nodes and to acquire tree data and meteorological data within the meteorological grid. The probability prediction module 202 is used to determine the probability of falling trees based on the tree data and the suspension data of the connection lines of the equipment nodes in the meteorological grid, according to the meteorological data and the falling tree database. Matching module 203 is used to match the dumping risk probability and the meteorological data with a preset rule base to determine the hazard factors; The prediction module 204 is used to predict the operating status of the power grid based on the risk factors and preset mapping relationships.

[0050] In one possible implementation, such as Figure 3 As shown, this application embodiment provides a terminal device 300, including: a memory 310, a processor 320, and a first computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the first computer program 311, it implements the steps of a power grid operation state prediction method.

[0051] In one possible implementation, such as Figure 4 As shown, this application embodiment provides a computer-readable storage medium 400, on which a second computer program 411 is stored. When the second computer program 411 is executed by a processor, it implements the steps of a power grid operation state prediction method.

[0052] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0053] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. 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 Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0054] 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 technical solutions 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 predicting the operating status of a power grid, characterized in that, include: Establish a meteorological grid centered on device nodes, and acquire tree data and meteorological data within the meteorological grid; Based on the tree data and the suspension data of the connection lines of the equipment nodes in the meteorological grid, the probability of falling risk is determined by the falling tree database according to the meteorological data; The probability of dumping risk and the meteorological data are matched with a preset rule base to determine the risk factors; The prediction of the power grid's operating status is completed based on the risk factors and preset mapping relationships.

2. The power grid operation status prediction method as described in claim 1, characterized in that, The steps of establishing a meteorological grid centered on device nodes and acquiring tree data and meteorological data within the meteorological grid include: Obtain the geographic coordinates of each device node within the target area; The meteorological data within the target area is divided into grids to obtain a meteorological grid; The meteorological grid where the device node is located is determined based on the geographic coordinates, and the meteorological grid where the device node is located is used as the target meteorological grid. The meteorological data corresponding to the target meteorological grid is recorded. Identify trees within the target meteorological grid that are located within a preset range centered on the connection line of the device node, and determine the tree data of the trees, which includes tree species, tree diameter at breast height (DBH), tree crown shape, tree location, and tree height.

3. The power grid operation status prediction method as described in claim 2, characterized in that, The step of determining the probability of falling trees based on the tree data and the suspension data of the connection lines of the equipment nodes in the meteorological grid, and using the meteorological data to determine the probability of falling trees through the fallen tree database, includes: Based on the tree location and tree height, and in conjunction with the suspension data of the connecting lines, the tree threat value is determined, and the tree hazard factor is determined based on the tree threat value; Based on the tree species category, the tree diameter at breast height (DBH) and the tree crown shape data, the probability of tree falling is determined by the fallen tree database according to the meteorological data, and the product of the tree hazard factor and the tree falling probability is recorded as the falling risk probability.

4. The power grid operation status prediction method as described in claim 3, characterized in that, The step of determining a tree threat value based on the tree's location and height, combined with the suspension data of the connecting line, and determining a tree hazard factor based on the tree threat value, includes: Based on the tree location and the suspension data, the horizontal distance between the tree and the connecting line is calculated, and the difference between the horizontal distance and the tree height is calculated to obtain the horizontal threat value; Determine whether the level of threat is below a first threshold; If the horizontal threat value is lower than the first threshold, the difference between the line suspension height and the tree height is calculated based on the tree height and the suspension data to obtain the height threat value; Determine whether the high threat value is lower than the second threshold; If the height threat value is lower than the second threshold, the sum of the height threat value and the horizontal threat value is calculated to obtain the total threat value, and the ratio of the height threat value to the total threat value is used as the tree risk factor.

5. The power grid operation status prediction method as described in claim 3, characterized in that, The step of determining the probability of tree falling based on the tree species category, the tree diameter at breast height (DBH) and the tree crown shape data, and using the fallen tree database according to the meteorological data, includes: Cluster analysis was performed on the fallen tree data in the fallen tree database according to the wind level data in the meteorological data to obtain multiple sets of fallen tree characteristics; The tree species category, tree diameter at breast height (DBH), and tree crown shape data are compared with multiple sets of features of fallen trees using Gower similarity calculation to generate a first similarity score. The first similarity scores are then sorted in descending order, and the first similarity score with the largest value is selected as the probability of tree falling.

6. The power grid operation status prediction method as described in claim 2, characterized in that, The step of identifying trees within the target meteorological grid that are located within a preset range centered on the connection line of the device node, and determining the tree data of the trees, includes: Obtain point cloud data of trees within a preset range with the connection line of the device node as the center line within the target meteorological grid, and determine the incomplete data of the trees; Using tree species, diameter at breast height (DBH), height, and crown shape as features, complete data that is closest to the incomplete data is selected from the historical database and used as template data. The template data and the incomplete data are horizontally sliced ​​at the same height to obtain data slices, and the perimeter points of each data slice are counted to obtain point count statistics. Based on the point count statistics, the data slices are sorted in ascending order by the count difference. The data slice with the smallest difference is selected as the baseline layer. The template data is scaled proportionally in the horizontal and vertical directions according to the baseline layer so that the perimeter points of the template data and the incomplete data are completely consistent at the baseline layer. At the edge of the missing region of the incomplete data, take three adjacent points in a clockwise direction to form a triangle. Calculate the displacement vector from the center of the triangle to the corresponding position of the template data. Apply the displacement vector to the corresponding point of the template data to make the corresponding region of the triangle translate to the missing region of the incomplete data until the missing region is completely covered by the continuous surface formed by the triangle, and obtain the completed incomplete data. The incomplete data is sliced ​​using contour lines, and the number of perimeter points in each slice is counted. The percentage difference is calculated based on the number of perimeter points in the incomplete data slices without missing points. If the percentage difference exceeds a preset threshold, the process continues with the step of acquiring point cloud data of trees within a preset range centered on the connection line of the device node within the target meteorological grid, and determining the incomplete data of the trees.

7. The power grid operation status prediction method as described in claim 1, characterized in that, The step of matching the dumping risk probability and the meteorological data with a preset rule base to determine the hazard factors includes: The historical number of faults is counted in the historical database based on the meteorological data, and the selected value is determined based on the historical number of faults. The probability of tipping over and the meteorological data are matched with a preset rule base to determine the initial hazard factor. The judgment rules in the rule base include: if the wind speed is greater than 8 m / s and the probability of tipping over is greater than 60, the risk factor increases by 0.3; if the humidity is greater than 80% and the temperature is greater than 30℃, the risk factor increases by 0.

2. Based on the selected value, the initial hazard factor is corrected to obtain the first hazard factor among the hazard factors; The device node whose first hazard factor is greater than or equal to the third threshold is taken as the target node, and the device node directly adjacent to the target node is taken as the neighbor node. The number of cascading failures and the total number of failures are determined based on historical databases, and the ratio of the number of cascading failures to the total number of failures is used as the diffusion coefficient. The risk factors of adjacent nodes are calculated based on the diffusion coefficient, and the risk factors of adjacent nodes are used as the second risk factor among the risk factors.

8. A power grid operation status prediction device, characterized in that, The device includes: The data acquisition module is used to establish a meteorological grid centered on the device nodes and to acquire tree data and meteorological data within the meteorological grid. The probability prediction module is used to determine the probability of falling trees based on the tree data and the suspension data of the connection lines of the equipment nodes in the meteorological grid, and to determine the probability of falling trees through the falling tree database according to the meteorological data. The matching module is used to match the dumping risk probability and the meteorological data with a preset rule base to determine the hazard factors; The prediction module is used to predict the operating status of the power grid based on hazard factors and preset mapping relationships.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the power grid operation state prediction method as described in any one of claims 1 to 4.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the power grid operation state prediction method as described in any one of claims 1 to 4.

Images