Power grid flexible inspection path planning method and system based on heuristic RRT algorithm

Abstract

The present invention provides a method and system for planning a flexible inspection path for a power grid based on a heuristic RRT algorithm, and relates to the field of flexible inspection for power grids. The specific scheme is as follows: the inspection information of a data center and a drone route library is collected to obtain original route data and environmental parameters; based on the requirements of power grid transmission services, the inspection position of a single pole tower is used as a basic unit, and the original route data is fragmented based on the environmental parameters to obtain waypoint data, and stored in a route database; based on the route database, a route reorganization model is established with the number of inspection positions of all preset pole towers and the inspection efficiency covered as parameters to be optimized; the route reorganization model is solved based on a heuristic RRT algorithm, and a flexible inspection path report for a power grid covering the preset pole tower inspection positions is generated and output. The present invention aims to improve the problem that the existing inspection path planning method is difficult to handle the complex environmental factors around the inspection object or the calculation results are not flexible enough.

Description

Technical Field

[0001] The present invention relates to the field of flexible inspection of power grids, and in particular to a method and system for planning a flexible inspection path for power grids based on a heuristic RRT algorithm. Background Art

[0002] Flexible power grid inspection refers to the regular, planned, and organized inspection, monitoring, evaluation, and maintenance of power grid systems and equipment. Compared with traditional inspections, it places more emphasis on the flexibility and intelligence of inspections. The purpose of flexible inspections is to promptly discover and eliminate hidden dangers of power grid equipment failures to ensure the safe operation of the power grid system and reliable power supply.

[0003] Flexible inspection route planning for power grids is an important basis for large-scale inspection of power facilities to improve inspection efficiency, ensure inspection safety, and quickly respond to inspection needs. The results of inspection route planning can be used to determine or guide the inspection route planning of various inspection methods such as drone inspections and manual inspections in various power professional fields such as transmission, distribution, and substation.

[0004] In addition, the flexible inspection path planning of power grids can also be used for the planning of adaptive inspections of UAVs of power facilities. For example, by adjusting the path finding algorithm to fly to the target point in the adaptive inspection, the convergence speed of the path optimization of the UAV to the target point can be planned, the stagnation time of the UAV can be reduced, the power loss can be reduced, and the inspection efficiency can be improved. In general, the flexible inspection path planning of power grids is one of the core ideas for improving the quality of large-scale inspections of power facilities. It provides important support for the safe, stable and efficient operation of power facility inspections.

[0005] With the continuous expansion of the scale of power grids and the increasing complexity of the network topology of power facilities, flexible inspection path planning of power grids is playing an increasingly important role in the large-scale inspection application of power systems. However, the flexible inspection path planning algorithms in the existing technologies either cannot effectively handle the complex environmental factors around the inspection objects of the power grids, or cannot meet the flexible power grid inspection business needs, resulting in insufficient applicability of the overall calculation results, or too long calculation time, affecting the overall inspection efficiency.

[0006] Therefore, it is urgent to develop a solution to solve the above problems. Summary of the invention

[0007] The purpose of the present invention is to provide a method and system for flexible inspection path planning of a power grid based on a heuristic RRT algorithm, aiming to improve the problem that the existing inspection path planning method is difficult to handle the complex environmental factors around the inspection object or the calculation results are not flexible enough.

[0008] The present invention provides a method and system for planning a flexible inspection path for a power grid based on a heuristic RRT algorithm, which adopts the following technical solutions:

[0009] In the first aspect, a flexible inspection path planning method for a power grid based on a heuristic RRT algorithm specifically includes: collecting inspection information from a data center and a drone route library to obtain original route data and environmental parameters;

[0010] Based on the power transmission business requirements of the power grid, the patrol position of a single tower is taken as the basic unit, and the original route data is fragmented based on the environmental parameters to obtain waypoint data, and stored in the route database;

[0011] Based on the route database, a route reorganization model is established with the number of inspection locations of all preset towers and inspection efficiency as the parameters to be optimized;

[0012] The route reorganization model is solved based on the heuristic RRT algorithm to generate and output a power grid flexible inspection path report covering the preset tower inspection locations.

[0013] The beneficial effect of the flexible inspection path planning method for power grids based on the heuristic RRT algorithm provided by the present invention is that by fragmenting the existing routes and combining the inspection environment data, according to the power grid inspection needs, a drone inspection route reorganization model with the goal of being more efficient, more flexible and safer is constructed. The reorganization model processes various special situations in different scenarios, and proposes different methods for different scenarios. The tower vertex method and the point filling method are proposed respectively. At the same time, the RRT algorithm is optimized by combining the existing point cloud, route, airspace and other spatial data to obtain the heuristic RRT algorithm for special scenarios. The three methods are combined to generate a flexible inspection path planning algorithm with wide flexibility, good safety and high accuracy. It also enhances the obstacle avoidance capability for complex power facility environmental factors, and provides strong support for the safety, stability and efficiency of power facility inspections.

[0014] Optionally, the inspection information of the data center and the drone route library is collected to obtain the original route data and environmental parameters, including:

[0015] The environmental parameters are environmental airspace parameters;

[0016] The environmental airspace parameters include an obstacle point set, a safety distance vector, and a safety area point set;

[0017] Among them, the obstacle point set and the safety distance vector are given input parameters;

[0018] The safety area point set is determined by calculating the obstacle point set and the safety distance vector.

[0019] Optionally, fragmenting the original route data based on the environmental parameters to obtain waypoint data, and storing the waypoint data in a route database, includes:

[0020] Based on the route results output by the existing route planning software and the power grid patrol business, the non-obstacle avoidance points of each route are marked as target points;

[0021] Wherein, the target point mark includes adding target point type and target point name attributes to the existing waypoint attributes;

[0022] The waypoints are grouped with the tower as the target to obtain a waypoint set, and the waypoint set is sorted based on the patrol location and priority. At the same time, the waypoint set setting is optimized, redundant points are removed, and obstacle avoidance and safe distance stop points are added;

[0023] Determine the point in the route where the waypoint rises vertically to a safe height and mark the safety point;

[0024] Based on the collected waypoint set, the altitude of the waypoints in the tower is calculated, and the tower vertex is obtained by combining the altitude of the waypoints of the current tower and the safety point;

[0025] The target point mark, waypoint set, safety point mark and tower vertex are uniformly processed to obtain waypoint data, which are classified and stored in a route database in a unified format.

[0026] Optionally, based on the route database, a route reorganization model is established with the number of inspection locations of all preset towers and the inspection efficiency as parameters to be optimized, including:

[0027] The safe point scenario is processed based on the point filling method, the unsafe point scenario is processed based on the heuristic RRT algorithm, the cross-tower scenario is processed based on the tower vertex method, the path efficiency is verified based on the simulated annealing algorithm, and the route reorganization model is constructed by integration;

[0028] The safety point scenario specifically refers to the current patrol target waypoint and the next patrol target waypoint being both safety points;

[0029] The point filling method refers to a method for planning a path between safety points, filling points at the current patrol target waypoint and the next patrol target waypoint vertically upward to a safe height, and then connecting the four points;

[0030] The non-safe point scenario specifically refers to a scenario where there is at least one non-safe point between the current patrol target waypoint and the next patrol target waypoint, and there is no complete route between the two points, or there is a complete route but the connection step length exceeds the extremely short step length;

[0031] The cross-tower scenario refers to a scenario where the current patrol target point waypoint crosses the tower to the next patrol target point waypoint;

[0032] The simulated annealing algorithm is used to evaluate the difference between the new route and the old route of the route planning results of two target waypoints.

[0033] Optionally, the non-safe point scenario is processed based on the heuristic RRT algorithm, including:

[0034] Input of heuristic model parameters; the heuristic model parameters include starting point parameters, target point parameters, environmental airspace parameters and heuristic area parameters;

[0035] The target point parameters include a patrol area point set and a patrol target point, and the heuristic area parameters are a random sampling point set;

[0036] The random sampling point set needs to be determined by calculating the obstacle point set and the safety distance vector;

[0037] The safety area is obtained by subtracting the non-navigable area from the patrol area and the total safety area is obtained by summing the two. The non-navigable area is determined by the safety distance diffused outward from the obstacle point set.

[0038] The first layer of heuristic area is obtained by subtracting the safe distance of the target point from the total safe area; the intersection of the first layer of heuristic area minus the safe distance and the total safe area is recursively obtained. Layer inspiration area;

[0039] The inverse transform sampling method is used to randomly sample the inspiration area to obtain a path that conforms to a specific probability distribution.

[0040] Optionally, solving the route reorganization model to generate and output a power grid flexible inspection route report covering a preset tower inspection location includes:

[0041] Classifying the target points according to the pole towers to construct a first array;

[0042] Set the starting point and the target point based on the first array;

[0043] The first array is traversed, and the starting point and the target point are updated based on the judgment of crossing the tower. The point-filling method is used to connect and add to the patrol target point set based on the safety point judgment. The current starting point and the target point are connected based on the heuristic RRT algorithm and the path result is verified until the first array is traversed. The verified path result is smoothed and output to obtain the power grid flexible inspection path report.

[0044] Optionally, the target points are classified according to the towers to construct a first array, and the starting point and the target point are set based on the first array, including:

[0045] Classify the target points by tower and construct the first array with index starting from 0;

[0046] The safety point is judged for the element No. 0 of the first array. When it is judged as a safety point, the first tower vertex is taken as the starting point and the element No. 0 of the first array is taken as the target point. Otherwise, the element No. 0 of the first array is taken as the starting point and the element No. 1 of the first array is taken as the target point.

[0047] Optionally, when there are towers between the starting point and the target point, the tower vertex method is used to add the tower vertices of two towers between the starting point and the target point to the patrol target point, and the starting point and the target point are reset.

[0048] Optionally, the verified path result is subjected to path smoothing processing and then outputted to obtain a power grid flexible inspection path report, including:

[0049] Construct a second array with a starting index of 0 based on the verified path result; wherein the elements in the second array are point coordinates;

[0050] Traverse the array from element No. 1, take the elements before and after the current element to get two points, calculate the angle between the two points and the current element, and determine whether the angle is a flat angle. If it is a flat angle, remove the current element.

[0051] In the second aspect, a power grid flexible inspection path planning system based on a heuristic RRT algorithm specifically includes: a data acquisition module for collecting inspection information from a data center and a drone route library to obtain original route data and environmental parameters;

[0052] A data processing module, based on the power transmission business requirements of the power grid, takes the patrol position of a single tower as a basic unit, fragments the original route data based on the environmental parameters to obtain waypoint data, and stores it in a route database;

[0053] The model building module, based on the route database, establishes a route reorganization model with the number of inspection locations of all preset towers and inspection efficiency as the parameters to be optimized;

[0054] The path generation module solves the route reorganization model based on the heuristic RRT algorithm, generates and outputs a power grid flexible inspection path report covering the preset tower inspection locations.

[0055] The beneficial effects of the second aspect can refer to the description of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A schematic diagram of the overall steps of a power grid flexible inspection path planning method based on a heuristic RRT algorithm provided by the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be understood by people with general skills in the field to which the present invention belongs. "Including" and similar words used in this article mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0058] The embodiment of the present invention provides a method for planning a flexible inspection path for a power grid based on a heuristic RRT algorithm, comprising the following steps:

[0059] S1, collect the inspection information of the data center and the drone route library to obtain the original route data and environmental parameters;

[0060] S2. Based on the power transmission business requirements of the power grid, taking the patrol position of a single tower as the basic unit, fragmenting the original route data based on the environmental parameters to obtain waypoint data, and storing it in the route database;

[0061] S3. Based on the route database, a route reorganization model is established with the number of inspection locations of all preset towers and inspection efficiency covered as parameters to be optimized;

[0062] S4. Solve the route reorganization model based on the heuristic RRT algorithm, generate and output a power grid flexible inspection path report covering the preset tower inspection locations.

[0063] In some embodiments, when executing step S1, the inspection information of the data center and the drone route library is collected to obtain the original route data and environmental parameters, including:

[0064] The environmental parameters are environmental airspace parameters;

[0065] The environmental airspace parameters include an obstacle point set, a safety distance vector, and a safety area point set;

[0066] Among them, the obstacle point set and the safety distance vector are given input parameters;

[0067] The safety area point set is determined by calculating the obstacle point set and the safety distance vector.

[0068] Specifically, the original route data includes the coordinates of the starting point and the end point of the route, the flight altitude, the speed, and the connection relationship between the waypoints.

[0069] Furthermore, the drone route library is a database that stores a variety of route information. Through the drone route library management system or dedicated software, query and extract the route data that meets the requirements, enter specific query conditions, such as route type, flight area, flight time, etc., to obtain the original route data.

[0070] In some embodiments, when executing step S2, the original route data is fragmented based on the environmental parameters to obtain waypoint data, and stored in the route database, specifically including:

[0071] S2-1. Based on the route results output by the existing route planning software and according to the power grid patrol business, mark the non-obstacle avoidance points of each route as target points;

[0072] S2-2, taking the tower as the target, the waypoints are grouped to obtain a waypoint set, and the waypoint set is sorted based on the patrol location and priority. At the same time, the waypoint set setting is optimized, redundant points are removed, and obstacle avoidance and safe distance stop points are added;

[0073] S2-3, determine the point in the route where the waypoint rises vertically to a safe height and mark the safety point;

[0074] S2-4, based on the collected waypoint set, calculate the altitude of the waypoints in the tower, and obtain the tower vertex by combining the altitude of the waypoints of the current tower and the safety point;

[0075] S2-5, uniformly processing the target point mark, waypoint set, safety point mark and tower vertex to obtain waypoint data, and storing them in a route database in a unified format.

[0076] Specifically, when executing step S2-1, the target point marking includes adding target point type and target point name attributes to the existing waypoint attributes.

[0077] Specifically, when executing step S2-2, the waypoint aggregation refers to associating the waypoint with the tower and establishing an association relationship between the waypoint and the tower. The aggregation method is to calculate the distance between the waypoint and the center point of each tower. The tower closest to the waypoint is the tower to which the waypoint belongs.

[0078] Furthermore, the distance calculation formula between the waypoint and the tower is as follows:

[0079] ;

[0080] In the formula, is the latitude and longitude coordinates of the waypoint, are the latitude and longitude coordinates of the tower.

[0081] It should be noted that waypoints refer to key locations on the flight path of the drone output by the route planning software, which are used for specific tasks such as target shooting. To ensure the effectiveness and pertinence of the inspection path, all waypoints in the original route data need to be deeply analyzed, accurately associated with the corresponding power towers, and grouped by tower. In this process, the waypoints are sorted according to the inspection location and priority, and the waypoint settings are optimized, redundant points are removed, and necessary obstacle avoidance and safe distance stops are added.

[0082] Specifically, when executing step S2-3, the safe height calculation method is:

[0083] ;

[0084] In the formula, is the tower elevation, Fixed height, It is a safe height;

[0085] Furthermore, the vertical distance calculation formula for each obstacle point to a waypoint is as follows:

[0086] ;

[0087] In the formula, is the altitude of the waypoint, is the altitude of the obstacle point, is the vertical distance;

[0088] For each obstacle, if the vertical distance is greater than the safe height, it is considered that there is no obstacle directly above the waypoint, and this waypoint is marked as a safe point.

[0089] Specifically, when executing step S2-4, the tower vertex refers to the waypoint set collected in S2-2, and the point that is higher than the safety height of the tower is recorded as the tower vertex. In the route planning, it is required that the first point of the patrol route of each tower must be higher than the safety height. Therefore, after processing according to the step S2-2, the tower with associated waypoints must have a tower vertex, that is, the point with the highest altitude of the waypoints associated with this tower.

[0090] Furthermore, the coordinates of the tower vertex are:

[0091] ;

[0092] In the formula, and are the longitude and latitude of the center point of the tower, For safe height.

[0093] It should be noted that in power inspections, especially in transmission and distribution line tower inspection scenarios, inspection tasks must be issued on a tower basis. At the same time, it is assumed that there are no obstacles directly above the tower. Therefore, in order to ensure the safety of inspections across towers, inspections across towers must first rise to the safe height of the tower before flying to the next tower.

[0094] Specifically, when executing step S2-5, the unified processing includes integrating the data to ensure the integrity and consistency of the data, cleaning the data to remove duplicate, invalid or erroneous data, and ensuring the accuracy and reliability of the data.

[0095] In some embodiments, when executing step S3, based on the route database, a route reorganization model is established with the number of all preset tower inspection locations and inspection efficiency covered as parameters to be optimized, including:

[0096] S3-1, processing safety point scenarios based on the point filling method;

[0097] S3-2, processing non-safe point scenarios based on heuristic RRT algorithm;

[0098] S3-3, processing cross-tower scenarios based on the tower vertex method;

[0099] S3-4. Verify the path efficiency based on the simulated annealing algorithm and integrate it to build a route reorganization model.

[0100] Specifically, when executing step S3-1, the safety point scenario specifically refers to that both the current patrol target waypoint and the next patrol target waypoint are safety points;

[0101] In fact, the point filling method refers to the path planning between the safety points, which is to fill points vertically upward to the safety height at the current patrol target waypoint and the next patrol target waypoint, and then connect the four points. For example, the current target waypoint is recorded as , the next target waypoint is recorded as , then they are vertically upwards to fill in points at a safe height, respectively. , , then connect Create a new patrol path.

[0102] In fact, when executing step S3-2, the non-safe point scene is processed based on the heuristic RRT algorithm, including:

[0103] Input of heuristic model parameters; the heuristic model parameters include starting point parameters, target point parameters, environmental airspace parameters and heuristic area parameters;

[0104] The target point parameters include a patrol area point set and a patrol target point, and the heuristic area parameters are a random sampling point set;

[0105] The random sampling point set needs to be determined by calculating the obstacle point set and the safety distance vector;

[0106] The safety area is obtained by subtracting the non-navigable area from the patrol area and the total safety area is obtained by summing the two. The non-navigable area is determined by the safety distance diffused outward from the obstacle point set.

[0107] Furthermore, the safe area is obtained by subtracting the non-airworthy area from the patrol area, and the calculation formula is as follows:

[0108] ;

[0109] In the formula, To patrol the area, It is a non-airworthy area;

[0110] Since different obstacle areas will set different safety distances, the final total safety area is:

[0111] ;

[0112] In the formula, To patrol the area, It is a non-airworthy area. is the total safety area, is the total number of samples, is the serial number;

[0113] The first layer of heuristic area is obtained by subtracting the safe distance of the target point from the total safe area; the intersection of the first layer of heuristic area minus the safe distance and the total safe area is recursively obtained. Layer inspiration area;

[0114] The inverse transform sampling method is used to randomly sample the inspiration area to obtain a path that conforms to a specific probability distribution.

[0115] Furthermore, the non-safe point scenario specifically refers to a scenario in which there is at least one non-safe point between the current patrol target waypoint and the next patrol target waypoint, and there is no complete route between the two points, or there is a complete route but the connection step length exceeds the extremely short step length.

[0116] Specifically, when executing step S3-3, the cross-tower scene refers to a scene where the current patrol target point waypoint crosses the tower to the next patrol target point waypoint;

[0117] In fact, to ensure patrol safety, for this scenario, it is necessary to connect from the current patrol target waypoint to the pole vertex of this pole first, then connect from the pole vertex of this pole to the pole vertex of the next pole, and finally connect from the pole vertex of the next pole to the next target point. In this scenario, the pole vertex is also a safe point, from the current patrol target waypoint to the current pole vertex and from the pole vertex of the next pole to the next patrol target waypoint.

[0118] Specifically, when executing step S3-4, the simulated annealing algorithm is used to evaluate the difference between the new route and the old route of the route planning results of the two target waypoints, integrating steps S3-1, S3-2, S3-3 and the processing steps of the simulated annealing algorithm and encapsulating them to obtain a route reorganization model.

[0119] In fact, for the two route planning results, there is a function difference :

[0120] ;

[0121] The new route planning result for the two destination waypoint route planning, The current route planning results for its route planning;

[0122] Furthermore, when When , the newly planned route is more efficient; when , it means that the new route planning has lower efficiency, but the route is still likely to be accepted. At this time, the acceptance probability calculation formula for the new route planning result is as follows:

[0123] ;

[0124] In the formula, is a constant parameter used to adjust the probability of accepting inefficient routes. is the probability of acceptance.

[0125] In some embodiments, when executing step S4, solving the route reorganization model, generating and outputting a power grid flexible inspection path report covering a preset tower inspection location, includes:

[0126] S4-1, classifying the target points according to the poles and towers to construct a first array, and setting the starting point and the target point based on the first array;

[0127] S4-2, traverse the first array, update the starting point and the target point based on the cross-tower judgment, connect them by the point-filling method based on the safety point judgment and add them to the patrol target point set, connect the current starting point and the target point based on the heuristic RRT algorithm and verify the path result until the first array is traversed, and smooth the verified path result and output it to obtain the power grid flexible inspection path report.

[0128] Specifically, when executing step S4-1, the target points are classified according to the towers to construct a first array, and the starting point and the target point are set based on the first array, including:

[0129] Classify the target points by tower and construct the first array with index starting from 0;

[0130] The safety point is judged for the element No. 0 of the first array. When it is judged as a safety point, the first tower vertex is taken as the starting point and the element No. 0 of the first array is taken as the target point. Otherwise, the element No. 0 of the first array is taken as the starting point and the element No. 1 of the first array is taken as the target point.

[0131] In fact, when there are towers between the starting point and the target point, the tower vertex method is used to add the tower vertices of two towers between the starting point and the target point to the patrol target point, and reset the starting point and the target point.

[0132] Specifically, when executing step S4-2, it includes:

[0133] S4-2-1: Set the starting point and target point;

[0134] S4-2-2: Determine whether the starting point to the target point crosses a tower. If so, add the tower vertices of two towers between the starting point and the target point to the patrol target point using the tower vertex method, and re-execute S4-2-1. Otherwise, execute S4-2-3.

[0135] S4-2-3: Determine whether the starting point and the target point are both tower vertices. If so, directly connect the starting point and the target point and execute S4-2-8. Otherwise, execute the next step S4-2-4.

[0136] S4-2-4: Determine whether the starting point and the target point are both safe points. If so, connect the starting point and the target point using the point-filling method and continue to execute S4-2-8. Otherwise, execute the next step S4-2-5.

[0137] S4-2-5: Start from the starting point to find out whether there is a safe point within the extremely short step length. If there is, add this safe point to the patrol target point and re-execute S4-2-1, otherwise execute the next step S4-2-6;

[0138] S4-2-6: Use heuristic RRT algorithm to connect the starting point and the target point;

[0139] S4-2-7: Use simulated annealing algorithm to check whether the RRT algorithm result is optimal. If it is, execute the next step S4-2-8, otherwise continue to execute S4-2-6;

[0140] S4-2-8: Determine whether the target point is the last point in the target set. If not, set the target point as the starting point, take the next point as the target point, and jump to S4-2-1. Otherwise, execute the next step S4-2-9.

[0141] S4-2-9: Path smoothing.

[0142] In fact, when executing step S4-2-9, the verified path result is subjected to path smoothing processing and then output to obtain a power grid flexible inspection path report, including:

[0143] Construct a second array with a starting index of 0 based on the verified path result; wherein the elements in the second array are point coordinates;

[0144] Traverse the array from the first element, take the elements before and after the current element to get two points, calculate the angle between the two points and the current element, and determine whether the angle is a flat angle. If it is a flat angle, remove the current element;

[0145] Furthermore, the angle formed by the two points and the current element is calculated, and the calculation formula is as follows:

[0146] ;

[0147] In the formula, , and is the point coordinate, is the angle.

[0148] Furthermore, we can determine the angle , if the following constraints are met, the current point can be eliminated:

[0149] ;

[0150] In the formula, is the critical value.

[0151] Although the embodiments of the present invention are described in detail above, it is obvious to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations are within the scope and spirit of the present invention as described in the claims. Moreover, the present invention described herein may have other embodiments and may be implemented or realized in a variety of ways.

Claims

1. A power grid flexible inspection path planning method based on heuristic RRT algorithm, characterized in that: include: Collect inspection information from the data center and the drone route library to obtain original route data and environmental parameters; Based on the power transmission business requirements of the power grid, the patrol position of a single tower is taken as the basic unit, and the original route data is fragmented based on the environmental parameters to obtain waypoint data, and stored in the route database; The safe point scenario is processed based on the point filling method, the unsafe point scenario is processed based on the heuristic RRT algorithm, the cross-tower scenario is processed based on the tower vertex method, the path efficiency is verified based on the simulated annealing algorithm, and the route reorganization model is constructed by integration; wherein, the unsafe point scenario is processed based on the heuristic RRT algorithm, including: the input of the heuristic model parameters; the heuristic model parameters include the starting point parameters, the target point parameters, the environmental airspace parameters and the heuristic area parameters; the target point parameters include the patrol area point set and the patrol target point, and the heuristic area parameters are the random sampling point set; wherein, the random sampling point set needs to be determined by the obstacle point set and the safety distance vector calculation; the safe area is obtained by subtracting the non-navigable area from the patrol area and the sum is obtained to obtain the total safe area, and the non-navigable area is determined by the safety distance diffused outward from the obstacle point set; the first-layer heuristic area is obtained by subtracting the safety distance diffused outward from the target point from the total safe area; the intersection of the first-layer heuristic area minus the safety distance and the total safe area is recursively obtained to obtain the first-layer heuristic area; the heuristic area is randomly sampled by the inverse transform sampling method to obtain a path that conforms to a specific probability distribution; The target points are classified by towers to construct a first array; the starting point and the target point are set based on the first array; the first array is traversed, the starting point and the target point are updated based on the cross-tower judgment, the point-filling method is used to connect and add to the patrol target point set based on the safety point judgment, the current starting point and the target point are connected based on the heuristic RRT algorithm and the path result is verified until the first array is traversed, and the verified path result is smoothed and output to obtain a power grid flexible inspection path report.

2. According to claim 1, a method for planning a flexible inspection path for a power grid based on a heuristic RRT algorithm is characterized in that: The inspection information of the data center and the drone route library is collected to obtain the original route data and environmental parameters, including: The environmental parameters are environmental airspace parameters; The environmental airspace parameters include an obstacle point set, a safety distance vector, and a safety area point set; Among them, the obstacle point set and the safety distance vector are given input parameters; The safety area point set is calculated from the obstacle point set and the safety distance vector.

3. The method for planning a flexible inspection path for a power grid based on a heuristic RRT algorithm according to claim 1, characterized in that: The original route data is fragmented based on the environmental parameters to obtain waypoint data, and stored in a route database, including: Based on the route results output by the existing route planning software and the power grid patrol business, the non-obstacle avoidance points of each route are marked as target points; Wherein, the target point mark includes adding target point type and target point name attributes to the existing waypoint attributes; The waypoints are grouped with the tower as the target to obtain a waypoint set, and the waypoint set is sorted based on the patrol location and priority. At the same time, the waypoint set setting is optimized, redundant points are removed, and obstacle avoidance and safe distance stop points are added; Determine the point in the route where the waypoint rises vertically to a safe height and mark the safety point; Based on the collected waypoint set, the altitude of the waypoints in the tower is calculated, and the tower vertex is obtained by combining the altitude of the waypoints of the current tower and the safety point; The target point mark, waypoint set, safety point mark and tower vertex are uniformly processed to obtain waypoint data, which are classified and stored in a route database in a unified format.

4. The method for planning a flexible inspection path for a power grid based on a heuristic RRT algorithm according to claim 1, characterized in that: The safety point scenario specifically refers to the current patrol target waypoint and the next patrol target waypoint being both safety points; The point filling method refers to a method for planning a path between safety points, filling points at the current patrol target waypoint and the next patrol target waypoint vertically upward to a safe height, and then connecting the four points; The non-safe point scenario specifically refers to a scenario where there is at least one non-safe point between the current patrol target waypoint and the next patrol target waypoint, and there is no complete route between the two points, or there is a complete route but the connection step length exceeds the extremely short step length; The cross-tower scenario refers to a scenario where the current patrol target point waypoint crosses the tower to the next patrol target point waypoint; The simulated annealing algorithm is used to evaluate the difference between the new route and the old route of the route planning results of two target waypoints.

5. According to claim 1, a method for planning a flexible inspection path for a power grid based on a heuristic RRT algorithm, characterized in that: Classifying the target points according to the towers to construct a first array, and setting the starting point and the target point based on the first array, including: Classify the target points by tower and construct the first array with index starting from 0; The safety point is judged for the element No. 0 of the first array. When it is judged as a safety point, the first tower vertex is taken as the starting point and the element No. 0 of the first array is taken as the target point. Otherwise, the element No. 0 of the first array is taken as the starting point and the element No. 1 of the first array is taken as the target point.

6. A method for planning a flexible inspection path for a power grid based on a heuristic RRT algorithm according to claim 5, characterized in that: When there are towers between the starting point and the target point, the tower vertex method is used to add the tower vertices of two towers between the starting point and the target point to the patrol target point, and reset the starting point and the target point.

7. The method for planning a flexible inspection path for a power grid based on a heuristic RRT algorithm according to claim 1, characterized in that: The verified path results are smoothed and then output to obtain a flexible inspection path report for the power grid, including: Construct a second array with a starting index of 0 based on the verified path result; wherein the elements in the second array are point coordinates; Traverse the array from element No. 1, take the elements before and after the current element to get two points, calculate the angle between the two points and the current element, and determine whether the angle is a flat angle. If it is a flat angle, remove the current element.

8. A power grid flexible inspection path planning system based on heuristic RRT algorithm, characterized in that: include: The data collection module is used to collect inspection information from the data center and the drone route library to obtain original route data and environmental parameters; A data processing module, based on the power transmission business requirements of the power grid, takes the patrol position of a single tower as a basic unit, fragments the original route data based on the environmental parameters to obtain waypoint data, and stores it in a route database; The model building module processes the safe point scenario based on the point filling method, processes the non-safe point scenario based on the heuristic RRT algorithm, processes the cross-tower scenario based on the tower vertex method, verifies the path efficiency based on the simulated annealing algorithm, and integrates and constructs the route reorganization model; wherein, the non-safe point scenario is processed based on the heuristic RRT algorithm, including: the input of the heuristic model parameters; the heuristic model parameters include the starting point parameters, the target point parameters, the environmental airspace parameters and the heuristic area parameters; the target point parameters include the patrol area point set and the patrol target point, and the heuristic area parameters are the random sampling point set; wherein, the random sampling point set needs to be determined by the obstacle point set and the safety distance vector calculation; the safe area is obtained by subtracting the non-navigable area from the patrol area and the total safety area is obtained by summing them up, and the non-navigable area is determined by the safety distance diffused outward from the obstacle point set; the first-layer heuristic area is obtained by subtracting the safety distance diffused outward from the target point from the total safety area; the intersection of the first-layer heuristic area minus the safety distance and the total safety area is recursively obtained to obtain the first-layer heuristic area; the heuristic area is randomly sampled by the inverse transform sampling method to obtain a path that conforms to a specific probability distribution; The path generation module classifies the target points according to the poles and towers to construct a first array; sets the starting point and the target point based on the first array; traverses the first array, updates the starting point and the target point based on the cross-tower judgment, connects them by the supplementary point method based on the safety point judgment and adds them to the patrol target point set, connects the current starting point and the target point based on the heuristic RRT algorithm and verifies the path result until the first array is traversed, and outputs the flexible power grid inspection path report after path smoothing processing on the verified path result.

Images