Unmanned aerial vehicle nest deployment and flight path joint optimization method for electric power inspection scene

Through the deployment of drone nests and joint optimization methods of flight trajectory, the drone inspection path is optimized using sparrow search and ant colony algorithm, which solves the problems of low efficiency and high cost of drones in power inspection, and achieves more efficient inspection and construction costs.

CN119002544APending Publication Date: 2024-11-22CHINA THREE GORGES UNIV
View PDF 0 Cites 5 Cited by

Patent Information

Application Number
CN202410894428.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-04
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

In the power inspection scenario, drone inspection efficiency is low and the coverage is limited, and the autonomous flight capacity is insufficient. It requires manual follow-up to complete tasks, resulting in high construction costs and long inspection time, and multiple drones coordinated inspections. Difficult to optimize.

Method used

Adopting the drone nest deployment and flight trajectory joint optimization method for power inspection scenarios, by establishing a drone inspection scenario model, setting constraints, mathematical modeling, and designing multi-UAV nest deployment and flight trajectory planning Algorithm, using the improved sparrow search K-Means algorithm and adaptive pheromone concentration ant colony algorithm, optimize the deployment and flight path of the drone nest, generate visual nest deployment and flight trajectory planning schemes to meet the flight time of the drone The longest constraint.

Benefits of technology

Minimize nesting costs and total inspection time, improve drone inspection efficiency, and reasonably build aircraft nests to meet actual needs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119002544A_ABST
    Figure CN119002544A_ABST
Patent Text Reader

Abstract

The invention discloses an unmanned aerial vehicle nest deployment and flight path joint optimization method for an electric power inspection scene, and aims at reducing the nesting cost and the total inspection time to the maximum extent by joint optimization of nesting and unmanned aerial vehicle paths and meeting the constraint condition that the flight time of an unmanned aerial vehicle is longest at the same time. Firstly, a problem is decomposed into a nest deployment sub-problem and an unmanned aerial vehicle path planning sub-problem, and then a joint optimization scheme is proposed, including a sparrow search K-mean algorithm for nest deployment and an adaptive pheromone concentration ant colony algorithm for unmanned aerial vehicle path planning. In order to simply verify the availability of the generated unmanned aerial vehicle nest deployment and unmanned aerial vehicle trajectory scheme, the generated scheme is visualized to check the final effect.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of UAV power inspection, and in particular relates to a method for jointly optimizing the deployment of UAV nests and flight trajectories for power inspection scenarios. Background Art

[0002] Drones have become one of the important equipment for power inspections, with their advantages that cannot be matched by manual inspections, such as strong mobility, high operating efficiency, and easy detection of defects on poles and towers. At present, drones are mainly used in different professional application scenarios such as transmission systems, substation systems, and distribution systems in power inspections. Among them, in transmission inspections, transmission lines span long distances and involve a wide area, mainly including high-voltage transmission lines, transmission towers, and insulators of transmission lines. In substation inspections, substations are usually located in urban suburbs or rural areas, occupying a large area and with a large number of equipment, requiring inspection and maintenance of various equipment. In distribution inspections, distribution lines usually cover various scenes such as cities, rural areas, and industrial areas, with different lengths and a wide coverage, requiring inspection and maintenance of lines, transformers, switchgear, etc. In the face of the urgent need for energy supply, it is crucial to periodically carry out power operation and inspection in the transmission, transformation, and distribution links. However, the wide distribution and large number of power equipment have put forward higher requirements for improving the efficiency of drone power inspections. On the one hand, the efficiency of single drone inspections is low and the coverage is limited. On the other hand, drones have insufficient autonomous flight capabilities and need to rely on manual following to gradually achieve full coverage of power inspection tasks.

[0003] In recent years, the deployment of drone nests has become a hot solution to the problem of insufficient drone endurance. The academic community has been exploring the best design, deployment strategy and optimization algorithm of drone nests through theoretical research and simulation experiments to improve the efficiency and coverage of power inspections. The industrial community has been actively exploring the actual application effect and economic benefits of drone nests in power inspections through field tests and application verification. At the same time, drone trajectory planning technology has also attracted attention from the fields of power, aerospace, intelligent transportation, geographic information systems, etc. By optimizing the flight path of drones, we can improve the efficiency of mission execution, reduce energy consumption costs, and ensure flight safety; however, drones still face the following challenges in autonomous intelligent inspection: (1) How to resolve the contradiction between the improvement of drone inspection efficiency and the increase in construction costs. Considering that when deploying drone nests, it is necessary not only to achieve full coverage of the search range, but also to balance the one-time investment in nest construction costs and the time cost of drone inspection; (2) How to solve the problem of coordinated inspection of multiple drones. Under the condition that the drone's own endurance is certain, the nest deployment and drone trajectory planning are jointly considered to determine the best solution; (3) For actual scenarios, how to reasonably select the algorithm to solve the problem is also crucial. Therefore, it is necessary to design a joint optimization method for drone nest deployment and flight trajectory for power inspection scenarios to solve the above problems. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a method for joint optimization of UAV nest deployment and flight trajectory for power inspection scenarios. The method aims to minimize the nesting cost and total inspection time by jointly optimizing nesting and UAV paths while satisfying the constraint of the longest UAV flight time.

[0005] In order to achieve the above technical effects, the technical solution adopted by the present invention is:

[0006] A method for joint optimization of drone nest deployment and flight trajectory for power inspection scenarios includes the following steps:

[0007] S1, establish drone inspection scene model;

[0008] S2, setting constraints for the drone inspection process;

[0009] S3, complete mathematical modeling according to the optimization goal;

[0010] S4, design multi-UAV nest deployment and flight trajectory planning algorithm;

[0011] S5, generate visual machine nest deployment and flight trajectory planning scheme;

[0012] S6, evaluate the performance of the proposed algorithm.

[0013] Preferably, in step S1, establishing a drone inspection scene model includes:

[0014] S101, Establish a multi-UAV transmission line tower inspection system:

[0015] The system includes U drones, N drone nests, and M towers to be inspected, which are represented as:

[0016]

[0017] Assuming that the flying height of the drone is constant during the inspection process, the drone flies in a two-dimensional Cartesian coordinate system. k The coordinates are M k =(x k ,y k );Machine nest N g The position coordinates are represented as N g =(x g ,y g );

[0018] S102, during the inspection process, the drone departs from the nest and performs hovering inspection according to the assigned task and the planned path; it is assumed that the take-off and landing time of the drone is negligible, and the drone travels at a constant speed V during the inspection process; the drone inspection is a hovering inspection, and T is used c Indicates that, in each round of inspection, it only hovers once at each tower, and the distance between two towers is:

[0019]

[0020] In the formula, and represents the horizontal and vertical coordinates of the jth tower, and represents the horizontal and vertical coordinates of the i-th tower;

[0021] S103, during the inspection process, when the UAV battery power is lower than the threshold, it returns to the nest to charge; a complete inspection process includes multiple round trips, represented as d f represents the fth round trip;

[0022] Set the total battery power of the drone to E, the flight power to P1, and the hovering power to P2; the remaining power is:

[0023] E remain =E-t1P1-t2P2;

[0024] Where t1 represents the flight time, and t2 represents the hovering time.

[0025] Preferably, in step S2, setting the constraint conditions of the drone inspection process includes:

[0026] Each machine nest is deployed on a pole tower, and each pole tower can deploy at most one machine nest;

[0027] Each inspection point must be visited once and only once;

[0028] The drone takes off from the nest and returns to the nest;

[0029] Each charging station is deployed at only one tower;

[0030] During the drone inspection process, the remaining power when returning to the machine nest shall not be less than 95% of the maximum power.

[0031] Preferably, in step S3, completing mathematical modeling according to the optimization target includes:

[0032] The cost of drone nest deployment and the total time of drone mission execution are normalized and jointly defined as the cost function minimization problem, which is expressed as:

[0033]

[0034] Among them, F = F1 + λ·F2 is the system cost objective function, and Represent the normalized time cost function and the nest construction cost function, T cycle is the inspection period of the task area, and T = t1 + t2 + t3 + t4 is the total inspection time, where Indicates the total time the drone flies between task points during the inspection process; is the path planning variable. If the UAV starts from the mission point M i Fly to mission point M j ,but otherwise Indicates the task point M i With the task point M j The distance between represents the total time spent by the drone hovering to check all mission points; T ci The time it takes for a drone to inspect a tower; represents the time it takes for the drone to travel back and forth between the machine nest and the mission point; y represents the charging variable. If the drone leaves the mission point M j Fly to the machine nest N i ,but otherwise z is the starting variable. If the UAV starts from the nest N i Fly to mission point M j ,but otherwise Indicates the total charging time of the drone during the inspection process; C total Indicates the overall cost of building the machine nest, Indicates the actual cost of building the machine nest; For deployment variables, if the machine nest N i Build at Mission Point M j If otherwise

[0035] Preferably, in step S4, designing a multi-UAV nest deployment and flight trajectory planning algorithm includes:

[0036] S401, cluster the towers using the improved sparrow search K-Means algorithm:

[0037] S4011, input tower point set And n cluster centers;

[0038] S4012, initialize the population, use the clustering criterion function as the fitness function, calculate the fitness and sort:

[0039]

[0040] Where n is the number of cluster centers; W j (t) represents the data sample in the jth cluster at the tth iteration; V j (t+1) Indicates the position of the jth cluster center in the t+1th time;

[0041] S4013, update the positions of the discoverer, follower and alerter in sequence:

[0042]

[0043] Where t is the current iteration number, X i,j is the position information of the i-th sparrow population in the j-th dimension, α is a random number between 0 and 1, iter max is the maximum number of iterations, Q is a random number that follows a normal distribution, L is a 1×d matrix with all elements set to 1, R2 is the safety value, ranging from 0 to 1, ST is the alarm value, ranging from 0.5 to 1; X P is the optimal position for the finder to search; X worst is the current global worst position; A is a 1×d matrix with random values ​​of 1 or -1, and satisfies A + =A T (AA T ) -1 ;Xbest The current global optimal value; β is the step size control parameter, which obeys the standard normal distribution; K1 is a random number between 0 and 1; f i is the current individual fitness value; f g is the current best fitness value; f g is the current worst fitness value; ε is a constant to avoid the denominator being 0;

[0044] S4014, repeat the above step S4013 until the iteration ends and n global optimal positions are obtained;

[0045] S4015, using n best positions as initial cluster centers, finding the cluster center closest to the tower, and dividing the tower points;

[0046] S4016, calculating new cluster centers and clustering criterion function values;

[0047] S4017, repeating steps S4015 and S4016 until the clustering criterion function value converges and the cluster center no longer changes, thereby achieving the division of the tower points;

[0048] S402, using adaptive pheromone concentration ant colony algorithm to perform path planning:

[0049] The pheromone concentration update rules are as follows:

[0050]

[0051] in Indicates that at time T+1, tower M i To tower M j pheromone value; Indicates the tower M at time T i To tower M j The pheromone value of t It is an adaptive pheromone volatilization factor; is the pheromone gain mode; Q is the pheromone intensity, which is a constant; L k is the length of the path currently traveled by ant k;

[0052] Adaptive pheromone volatility factor ρ t for:

[0053]

[0054] Where κ is the adjustment coefficient, which is an integer less than 1; L SG is the Euclidean distance from the starting point S to the target point G of the path; L k is the path length of the current iteration path of ant k;

[0055] The calculation formula of the heuristic function is as follows:

[0056]

[0057] in Indicates tower M i To tower M j The distance between Indicates tower M j The distance to the end point s; δ, ε are the corresponding weights;

[0058] The specific steps are:

[0059] S4021, obtaining the clustering result of S401;

[0060] S4022, initializing the parameters of the ant colony algorithm, including pheromone intensity Q and adjustment coefficient k;

[0061] S4023, selecting the next tower point according to the probability formula;

[0062] S4024, if all pole tower points have been traversed, calculate the total inspection time; otherwise, continue traversing until all pole tower points have been traversed;

[0063] S4025, update the pheromone according to the pheromone update formula, and determine whether the maximum number of iterations has been reached. If so, output the optimal path, otherwise repeat steps S4021 to S4024.

[0064] Preferably, in step S5, generating a visualized machine nest deployment and flight trajectory planning scheme includes:

[0065] The coordinates of the tower points are displayed in a two-dimensional Cartesian coordinate system. According to the drone nest site selection plan, the location of the nest is marked, and the towers are connected in sequence according to the path planning plan to form the final path planning map.

[0066] Preferably, in step S6, evaluating the performance of the proposed algorithm includes:

[0067] S601, obtaining the Cartesian coordinate points of the tower;

[0068] S602, setting parameters of the sparrow search algorithm;

[0069] S603, setting parameters of the ant colony algorithm;

[0070] S604, setting the drone's own performance parameters.

[0071] The beneficial effects of the present invention are as follows:

[0072] The present invention aims to resolve the contradiction between the improvement of drone inspection efficiency and the increase of construction cost. It establishes an objective function for jointly optimizing the nest deployment cost and the total inspection time, while satisfying the constraint of the longest flight time of the drone, thereby minimizing the nesting cost and increasing the total inspection time. It can be reasonably constructed according to actual engineering needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 Schematic diagram of drone inspection scenario. DETAILED DESCRIPTION

[0074] Embodiment 1:

[0075] like Figure 1 As shown, a method for joint optimization of drone nest deployment and flight trajectory for power inspection scenarios includes the following steps:

[0076] S1, establish drone inspection scene model;

[0077] S2, setting constraints for the drone inspection process;

[0078] S3, complete mathematical modeling according to the optimization goal;

[0079] S4, design multi-UAV nest deployment and flight trajectory planning algorithm;

[0080] S5, generate visual machine nest deployment and flight trajectory planning scheme;

[0081] S6, evaluate the performance of the proposed algorithm.

[0082] Preferably, in step S1, establishing a drone inspection scene model includes:

[0083] S101, Establish a multi-UAV transmission line tower inspection system:

[0084] The system includes U drones, N drone nests, and M towers to be inspected, which are represented as:

[0085]

[0086] Assuming that the flying height of the drone is constant during the inspection process, the drone flies in a two-dimensional Cartesian coordinate system. k The coordinates are M k =(x k ,y k );Machine nest N g The position coordinates are represented as N g =(x g ,y g );

[0087] S102, during the inspection process, the drone departs from the nest and performs hovering inspection according to the assigned task and the planned path; it is assumed that the take-off and landing time of the drone is negligible, and the drone travels at a constant speed V during the inspection process; the drone inspection is a hovering inspection, and T is used c Indicates that, in each round of inspection, it only hovers once at each tower, and the distance between the two towers is:

[0088]

[0089] In the formula, and represents the horizontal and vertical coordinates of the jth tower, and represents the horizontal and vertical coordinates of the i-th tower;

[0090] S103, during the inspection process, when the UAV battery power is lower than the threshold, it returns to the nest to charge; a complete inspection process includes multiple round trips, represented as d f represents the fth round trip;

[0091] Set the total battery power of the drone to E, the flight power to P1, and the hovering power to P2; the remaining power is:

[0092] E remain =E-t1P1-t2P2;

[0093] Where t1 represents the flight time, and t2 represents the hovering time.

[0094] Preferably, in step S2, setting the constraint conditions of the drone inspection process includes:

[0095] Each machine nest is deployed on a pole tower, and each pole tower can deploy at most one machine nest;

[0096] Each inspection point must be visited once and only once;

[0097] The drone takes off from the nest and returns to the nest;

[0098] Each charging station is deployed at only one tower;

[0099] During the drone inspection process, the remaining power when returning to the machine nest shall not be less than 95% of the maximum power.

[0100] Preferably, in step S3, completing mathematical modeling according to the optimization target includes:

[0101] The cost of drone nest deployment and the total time of drone mission execution are normalized and jointly defined as the cost function minimization problem, which is expressed as:

[0102]

[0103] Among them, F = F1 + λ·F2 is the system cost objective function, and Represent the normalized time cost function and the nest construction cost function, T cycle is the inspection period of the task area, and T = t1 + t2 + t3 + t4 is the total inspection time, where Indicates the total time the drone flies between task points during the inspection process; is the path planning variable. If the UAV starts from the mission point M i Fly to mission point M j ,but otherwise Indicates the task point M i With the task point M j The distance between represents the total time spent by the drone hovering to check all mission points; T ci The time it takes for a drone to inspect a tower; represents the time it takes for the drone to travel back and forth between the machine nest and the mission point; y represents the charging variable. If the drone leaves the mission point M j Fly to the machine nest N i ,but otherwise z is the starting variable. If the UAV starts from the nest N i Fly to mission point M j ,but otherwise Indicates the total charging time of the drone during the inspection process; C total Indicates the overall cost of building the machine nest, Indicates the actual cost of building the machine nest; For deployment variables, if the machine nest N i Build at Mission Point M j If otherwise

[0104] Preferably, in step S4, designing a multi-UAV nest deployment and flight trajectory planning algorithm includes:

[0105] S401, cluster the towers using the improved sparrow search K-Means algorithm:

[0106] S4011, input tower point set And n cluster centers;

[0107] S4012, initialize the population, use the clustering criterion function as the fitness function, calculate the fitness and sort:

[0108]

[0109] Where n is the number of cluster centers; W j (t) represents the data sample in the jth cluster at the tth iteration; V j (t+1) Indicates the position of the jth cluster center in the t+1th time;

[0110] S4013, update the positions of the discoverer, follower and alerter in sequence:

[0111]

[0112]

[0113] Where t is the current iteration number, X i,j is the position information of the i-th sparrow population in the j-th dimension, α is a random number between 0 and 1, iter max is the maximum number of iterations, Q is a random number that follows a normal distribution, L is a 1×d matrix with all elements set to 1, R2 is the safety value, ranging from 0 to 1, ST is the alarm value, ranging from 0.5 to 1; X P is the optimal position for the finder to search; X worst is the current global worst position; A is a 1×d matrix with random values ​​of 1 or -1, and satisfies A + =A T (AA T ) -1 ;X best The current global optimal value; β is the step size control parameter, which obeys the standard normal distribution; K1 is a random number between 0 and 1; f i is the current individual fitness value; f g is the current best fitness value; f g is the current worst fitness value; ε is a constant to avoid the denominator being 0;

[0114] S4014, repeat the above step S4013 until the iteration ends and n global optimal positions are obtained;

[0115] S4015, using n best positions as initial cluster centers, finding the cluster center closest to the tower, and dividing the tower points;

[0116] S4016, calculating new cluster centers and clustering criterion function values;

[0117] S4017, repeating steps S4015 and S4016 until the clustering criterion function value converges and the cluster center no longer changes, thereby achieving the division of the tower points;

[0118] S402, using adaptive pheromone concentration ant colony algorithm to perform path planning:

[0119] The pheromone concentration update rules are as follows:

[0120]

[0121] in Indicates that at time T+1, tower M i To tower M j pheromone value; Indicates the tower M at time T i To tower M j The pheromone value of t It is an adaptive pheromone volatilization factor; is the pheromone gain mode; Q is the pheromone intensity, which is a constant; L k is the length of the path currently traveled by ant k;

[0122] Adaptive pheromone volatility factor ρ t for:

[0123]

[0124] Where κ is the adjustment coefficient, which is an integer less than 1; L SG is the Euclidean distance from the starting point S to the target point G of the path; L k is the path length of the current iteration path of ant k;

[0125] The calculation formula of the heuristic function is as follows:

[0126]

[0127] in Indicates tower M i To tower M j The distance between Indicates tower M j The distance to the end point s; δ, ε are the corresponding weights;

[0128] The specific steps are:

[0129] S4021, obtaining the clustering result of S401;

[0130] S4022, initializing the parameters of the ant colony algorithm, including pheromone intensity Q and adjustment coefficient k;

[0131] S4023, selecting the next tower point according to the probability formula;

[0132] S4024, if all pole tower points have been traversed, calculate the total inspection time; otherwise, continue traversing until all pole tower points have been traversed;

[0133] S4025, update the pheromone according to the pheromone update formula, and determine whether the maximum number of iterations has been reached. If so, output the optimal path, otherwise repeat steps S4021 to S4024.

[0134] Preferably, in step S5, generating a visualized machine nest deployment and flight trajectory planning scheme includes:

[0135] The coordinates of the tower points are displayed in a two-dimensional Cartesian coordinate system. According to the drone nest site selection plan, the location of the nest is marked, and the towers are connected in sequence according to the path planning plan to form the final path planning map.

[0136] Preferably, in step S6, evaluating the performance of the proposed algorithm includes:

[0137] S601, obtaining the Cartesian coordinate points of the tower;

[0138] S602, setting parameters of the sparrow search algorithm;

[0139] S603, setting parameters of the ant colony algorithm;

[0140] S604, setting the drone's own performance parameters.

[0141] Embodiment 2:

[0142] In order to solve the contradiction between the improvement of drone inspection efficiency and the increase of construction cost, considering that when deploying drone nests, it is necessary not only to achieve full coverage of the search range, but also to weigh the one-time investment cost of nest construction and the time cost of drone inspection. Therefore, different weights are selected to compare the objective functions.

[0143] Table 1 shows the number of drone nests and the total inspection time optimized by the adaptive pheromone concentration ant colony algorithm and the traditional ant colony algorithm under different weight coefficients. When the weight coefficient increases over time, it can be observed that the number of drone nests decreases from 2 to 1, while the total inspection time increases. This is because different weight coefficients have different degrees of optimization for drone inspection time and nest cost. In particular, when the weight coefficient is small, more emphasis is placed on optimizing drone inspection time, while as the weight coefficient increases, the focus shifts to optimizing drone nesting costs.

[0144]

[0145] Table 1: Comparison of simulation results of traditional ant colony algorithm and adaptive pheromone concentration ant colony algorithm under different weight factors;

[0146] Table 2 is the pseudo code of the sparrow search clustering algorithm, which provides the specific process of the sparrow search clustering algorithm.

[0147]

[0148] Table 2: Sparrow search clustering algorithm;

[0149] Table 3 shows the pseudo code of the adaptive pheromone concentration ant colony algorithm and provides the specific process of the adaptive pheromone concentration ant colony algorithm.

[0150]

[0151]

[0152] Table 3: Adaptive pheromone concentration ant colony algorithm;

[0153] It can be seen from Table 1 that although increasing λ can reduce the construction cost of the drone nest, the total inspection time will increase accordingly.

Claims

1. A joint optimization method for drone nest deployment and flight trajectory for power inspection scenarios, characterized in that: The following steps are involved: S1, establish drone inspection scene model; S2, setting constraints for the drone inspection process; S3, complete mathematical modeling according to the optimization goal; S4, design multi-UAV nest deployment and flight trajectory planning algorithm; S5, generate visual machine nest deployment and flight trajectory planning scheme; S6, evaluate the performance of the proposed algorithm.

2. The method for joint optimization of drone nest deployment and flight trajectory for power inspection scenarios according to claim 1 is characterized in that: In step S1, establishing a drone inspection scene model includes: S101, Establish a multi-UAV transmission line tower inspection system: The system includes U drones, N drone nests, and M towers to be inspected, which are represented as: Assuming that the flying height of the drone is constant during the inspection process, the drone flies in a two-dimensional Cartesian coordinate system. k The coordinates are M k =(x k ,y k );Machine nest N g The position coordinates are represented as N g =(x g ,y g ); S102, during the inspection process, the drone departs from the machine nest and performs hovering inspection according to the assigned task and the planned path; the drone is set to travel at a constant speed V during the inspection process; the drone inspection is a hovering inspection, and T is used. c Indicates that, in each round of inspection, it only hovers once at each tower, and the distance between two towers is: In the formula, and represents the horizontal and vertical coordinates of the jth tower, and represents the horizontal and vertical coordinates of the i-th tower; S103, during the inspection process, when the UAV battery power is lower than the threshold, it returns to the nest to charge; a complete inspection process includes multiple round trips, represented as d f represents the fth round trip; Set the total battery power of the drone to E, the flight power to P1, and the hovering power to P2; the remaining power is: AND remain =E-t1P1-t2P2; Where t1 represents the flight time, and t2 represents the hovering time.

3. The method for joint optimization of drone nest deployment and flight trajectory for power inspection scenarios according to claim 1 is characterized in that: In step S2, setting the constraints of the drone inspection process includes: Each machine nest is deployed on a pole tower, and each pole tower can deploy at most one machine nest; Each inspection point must be visited once and only once; The drone takes off from the nest and returns to the nest; Each charging station is deployed at only one tower; During the drone inspection process, the remaining power when returning to the machine nest shall not be less than 95% of the maximum power.

4. The method for joint optimization of drone nest deployment and flight trajectory for power inspection scenarios according to claim 1 is characterized in that: In step S3, completing mathematical modeling according to the optimization target includes: The cost of drone nest deployment and the total time of drone mission execution are normalized and jointly defined as the cost function minimization problem, which is expressed as: Among them, F = F1 + λ·F2 is the system cost objective function, and Represent the normalized time cost function and the nest construction cost function, T cycle is the inspection period of the task area, and T = t1 + t2 + t3 + t4 is the total inspection time, where Indicates the total time the drone flies between task points during the inspection process; is the path planning variable. If the UAV starts from the mission point M i Fly to mission point M j ,but otherwise Indicates the task point M i With the task point M j The distance between represents the total time spent by the drone hovering to check all mission points; T ci The time it takes for a drone to inspect a tower; represents the time it takes for the drone to travel back and forth between the machine nest and the mission point; y represents the charging variable. If the drone leaves the mission point M j Fly to the machine nest N i ,but otherwise z is the starting variable. If the UAV starts from the nest N i Fly to mission point M j ,but otherwise Indicates the total charging time of the drone during the inspection process; C total Indicates the overall cost of building the machine nest, Indicates the actual cost of building the machine nest; For deployment variables, if the machine nest N i Build at Mission Point M j If otherwise 5. The method for joint optimization of drone nest deployment and flight trajectory for power inspection scenarios according to claim 1 is characterized in that: In step S4, designing a multi-UAV nest deployment and flight trajectory planning algorithm includes: S401, cluster the towers using the improved sparrow search K-Means algorithm: S4011, input tower point set And n cluster centers; S4012, initialize the population, use the clustering criterion function as the fitness function, calculate the fitness and sort: Where n is the number of cluster centers; W j (t) represents the data sample in the jth cluster at the tth iteration; V j (t+1) Indicates the position of the jth cluster center in the t+1th time; S4013, update the positions of the discoverer, follower and alerter in sequence: Where t is the current iteration number, X i,j is the position information of the i-th sparrow population in the j-th dimension, α is a random number between 0 and 1, iter max is the maximum number of iterations, Q is a random number that follows a normal distribution, L is a 1×d matrix with all elements set to 1, R2 is the safety value, ranging from 0 to 1, ST is the alarm value, ranging from 0.5 to 1; X P is the optimal position for the finder to search; X worst is the current global worst position; A is a 1×d matrix with random values ​​of 1 or -1, and satisfies A + =A T (AA T ) -1 ;X best The current global optimal value; β is the step size control parameter, which obeys the standard normal distribution; K1 is a random number between 0 and 1; f i is the current individual fitness value; f g is the current best fitness value; f g is the current worst fitness value; ε is a constant to avoid the denominator being 0; S4014, repeat the above step S4013 until the iteration ends and n global optimal positions are obtained; S4015, using n best positions as initial cluster centers, finding the cluster center closest to the tower, and dividing the tower points; S4016, calculating new cluster centers and clustering criterion function values; S4017, repeating steps S4015 and S4016 until the clustering criterion function value converges and the cluster center no longer changes, thereby achieving the division of the tower points; S402, using adaptive pheromone concentration ant colony algorithm to perform path planning: The pheromone concentration update rules are as follows: in Indicates that at time T+1, tower M i To tower M j pheromone value; Indicates the tower M at time T i To tower M j The pheromone value of t It is an adaptive pheromone volatilization factor; is the pheromone gain mode; Q is the pheromone intensity, which is a constant; L k is the length of the path currently traveled by ant k; Adaptive pheromone volatility factor ρ t for: Where κ is the adjustment coefficient, which is an integer less than 1; L SG is the Euclidean distance from the starting point S to the target point G of the path; L k is the path length of the current iteration path of ant k; The calculation formula of the heuristic function is as follows: in Indicates tower M i To tower M j The distance between Indicates tower M j The distance to the end point s; δ, ε are the corresponding weights; The specific steps are: S4021, obtaining the clustering result of S401; S4022, initializing the parameters of the ant colony algorithm, including pheromone intensity Q and adjustment coefficient k; S4023, selecting the next tower point according to the probability formula; S4024, if all pole tower points have been traversed, calculate the total inspection time; otherwise, continue traversing until all pole tower points have been traversed; S4025, update the pheromone according to the pheromone update formula, and determine whether the maximum number of iterations has been reached. If so, output the optimal path, otherwise repeat steps S4021 to S4024.

6. The method for joint optimization of drone nest deployment and flight trajectory for power inspection scenarios according to claim 1 is characterized in that: In step S5, generating a visualized nest deployment and flight trajectory planning scheme includes: The coordinates of the tower points are displayed in a two-dimensional Cartesian coordinate system. According to the drone nest site selection plan, the location of the nest is marked, and the towers are connected in sequence according to the path planning plan to form the final path planning map.

7. The method for joint optimization of drone nest deployment and flight trajectory for power inspection scenarios according to claim 1 is characterized in that: In step S6, evaluating the performance of the proposed algorithm includes: S601, obtaining the Cartesian coordinate points of the tower; S602, setting parameters of the sparrow search algorithm; S603, setting parameters of the ant colony algorithm; S604, setting the drone's own performance parameters.

Citation Information

Cited By

  • Unmanned aerial vehicle power distribution network autonomous inspection path planning method and system adaptive to unmanned aerial vehicle nest

    CN119620781A

  • An autonomous inspection path planning method and system for a drone power distribution network adapted to a drone nest

    CN119620781B

  • Unmanned aerial vehicle charging path planning method based on hybrid artificial bee colony and ant colony algorithm

    CN120406518A

  • Unmanned aerial vehicle nest space optimal configuration method and system for comprehensive inspection

    CN121146456A

  • A method and system for optimizing the configuration of UAV nest space for integrated inspection

    CN121146456B