CN-121995928-A - Heuristic and local search-based photovoltaic unmanned aerial vehicle cleaning path generation method

Abstract

The invention relates to a photovoltaic unmanned aerial vehicle cleaning path generation method based on heuristic and local search, which uses a photovoltaic cleaning path planning algorithm taking weighted delay as a main factor to generate an unmanned aerial vehicle cleaning candidate path considering three constraint targets of array priority, dirt weight priority and flight path short in a short time. In the invention, in the framework of 'first array and then assembly', the 'weight-distance' is used for evaluating pollution and controlling cost, the method can quickly converge to an engineering available scheme, lambda adopts the 'cluster weight and adjacent distance median' for normalization, different power station scales and layout densities are adapted, manual repeated parameter adjustment is not needed, disturbance restarting and minimum increment reinsertion are used for disturbing and rearranging partial paths, the diversity and quality of solutions are obviously improved, a stronger candidate path upper limit is provided for subsequent scoring, the whole process only depends on basic quantities such as cluster centroid coordinates, cluster weights and the like, the complexity is controllable, and the method is suitable for quickly generating a plurality of candidate paths.

Inventors

• Geng changxing
• XU XIWEN
• Liu Chuanghai
• JIA MINGYUAN
• WANG YONG

Assignees

• 苏州大学
• 飒沓机器人科技(苏州)有限公司

Dates

Publication Date

20260508

Application Date

20251205

Claims (3)

1. 1. A photovoltaic unmanned aerial vehicle cleaning path generation method based on heuristic and local search is characterized by comprising the following steps: Step S1, target detection and component set construction Completing component target detection on the input orthophoto map to obtain a component external rectangular set: step S2, constructing a dirt proportion and an array cluster Obtaining a dirt proportion r k in the assembly, and polymerizing the assembly into an array cluster according to an array polygon, wherein: The minimum particle unit corresponding to the single photovoltaic panel or the detection output is recorded as a kth component, the viscera pollution proportion r k E [0,1] of the component is detected, and the component is attached with a non-negative weight W module for reflecting the importance of the component to the cleaning decision; Generation of array cluster boundaries Cj: Detecting all the component frames by using YOLOv, and writing the whole component frames into a binary mask M on the whole graph; performing single expansion on M by using a rectangular kernel K to connect adjacent/neighbor components; Extracting an outer contour of an expansion result, and obtaining an array cluster boundary Cj by using a polygon approximation method RDP; Each cluster contains a plurality of photovoltaic modules and weights thereof: w module ＝Tier(r k )∈{1,2,3}; Step S3, path planning Step S3.1, constructing a solution, namely taking the heaviest cluster as a starting point, taking the current cluster as i, taking a candidate non-access cluster set as U, defining local cost for any j E U, namely, cost (j) =λd (i, j) -W j , namely, setting the smallest cost as the next point, sequencing the points, wherein d (i, j) is Euclidean distance between i and the mass centers of j array clusters, W j is the weight of the array clusters, W 0 is the median of cluster weights, d 0 is the median of common neighboring cluster spacings; Step S3.2, local search, 2-opt and repositioning relocate are carried out on a single route, and the best improvement is carried out at the cost: The relocation relocate is used as a local search operation, any non-starting point node in the path is removed from the original position and inserted into other positions of the same path on the premise of not changing the starting point, the incremental cost delta is used as an evaluation criterion, and the best improvement principle is adopted to select whether to accept the movement; A movable object, which is any node v except a starting point; Allowed insertion position, all insertion slits in the route, but not allowing v to be inserted into the first position; The goal is to calculate the incremental Cost delta for each candidate insertion location using a defined Cost formula, delta = Cost (new) -Cost (old), take the optimal path, perform the move if delta <0, repeat until there is no improvement, Step S3.3, disturbance restarting, namely randomly moving out non-starting point nodes of p, and then reinserting one by one at the cost of the minimum increment so as to jump out of local search to find a better solution; S3.4, circulating until the number of the iterations is reached or the upper limit of no improvement is reached, and outputting an optimal cluster sequence; s3.5, merging the produced cluster sequences and removing the weight; step S4, in-cluster origin definition The starting point of a component of the first array cluster is that a component with the highest dirt occupation ratio r k is selected in the cluster as the starting point; The component start point of the subsequent array cluster is that the exit component of the previous cluster is P out , and the following priority selection start points are selected in the current cluster: Selecting the nearest Euclidean distance with P out from the component set with weight=3; if the weight=3 is not available, the method is selected in the weight=2, and if the weight=1 is not available, the method is selected in the weight=1; If the parallel points appear, all the candidate paths are included; Step S5, in-cluster component access sequence and endpoint definition The kth component box in cluster c i Taking the left middle point/the right middle point as a cleaning end point: The lateral crossover cleaning time t k for a monolithic assembly is defined as: Wherein v clean is the component cleaning speed in pixels per second, t over is the fixed preparation/trigger overhead time per block of component in seconds; The initial order within the cluster is optionally one of three heuristics: 1) Decreasing according to w module ; 2) Nearest neighbor +2-opt; 3) greedy_weighted, maximizing greedy for W- β·d, where W is the weight of the component, d is the linear distance from the current point to the centroid of the component, β is the penalty factor for distance, the greater β is the more aggressive the closer β is to first choose, the smaller β is the more aggressive the first is to second choose; step S6, dynamic programming of intra-cluster direction selection The access sequence is p 1 ,p 2 ,...,p n , and each component board has two directions: 1) d=0 washes L k →R k from left to right, entry point S k (0)＝L k ; 2) d=1 right to left wash R k →L k , entry point S k (1)＝R k ; setting a span displacement speed v move , defining dp: dp [ k ] [ d ] = complete to kth block assembly and kth block cleaning direction is dmin integration time; the displacement time from the end point of the previous block to the start point of the current block is considered in recursion: In the formula, the initial value dp considers the distance between the flying spot and the cleaning starting point of the first block, and backtracks to obtain the optimal direction sequence of the whole cluster and the endpoint sequence entering/leaving on the kth plate.
2. 2. The method for generating the cleaning path of the photovoltaic unmanned aerial vehicle based on heuristic and local search according to claim 1, wherein in the step S2, C 1 is 5%, C 2 is 15%, C 3 is 30%, and components with the dirt proportion lower than the threshold r k ＜C 1 are skipped and do not participate in cleaning and planning.
3. 3. The method for generating the cleaning path of the photovoltaic unmanned aerial vehicle based on heuristic and local search according to claim 1, wherein the specific value method of d 0 in step S3.1 is as follows: cluster centroid list p= [ (x 1 ,y 1 ),...,(x N ,y N ) ]: 1) For each i: i.e. traversing all other clusters, and j+.i; 2) All d i were collected; 3) The median was taken as d 0 .

Description

Heuristic and local search-based photovoltaic unmanned aerial vehicle cleaning path generation method Technical Field The invention relates to the technical field of unmanned plane path optimization and intelligent scheduling, in particular to candidate path generation guided by a weighted minimum delay target in an array priority cleaning scene. Background Under the preferential cleaning constraint of the photovoltaic scene array, the simple nearest neighbor or shortest path algorithm often ignores the time sequence benefit of the pollution weight, and the purely pursuing the preferential cleaning mode of heavy pollution can cause the overlong navigation process. The traditional path planning algorithm is difficult to solve in real time under the condition of large-scale array cluster number, and a heuristic algorithm which takes weight priority and voyage inhibition as principles and can be continuously improved within a limited time is needed to generate candidate paths. In order to solve the problems, the invention provides a set of photovoltaic cleaning path planning algorithm taking weighted delay as a main factor, so as to generate an unmanned plane cleaning candidate path considering three constraint targets of array priority, dirt weight priority and flight path short in a short time. Disclosure of Invention The invention aims to overcome the problems in the prior art and provides a photovoltaic unmanned aerial vehicle cleaning path generation method based on heuristic and local search. In order to achieve the technical purpose and the technical effect, the invention is realized by the following technical scheme: A photovoltaic unmanned aerial vehicle cleaning path generation method based on heuristic and local search comprises the following steps: Step S1, target detection and component set construction Completing component target detection on the input orthophoto map to obtain a component external rectangular set: step S2, constructing a dirt proportion and an array cluster Obtaining a dirt proportion r k in the assembly, and polymerizing the assembly into an array cluster according to an array polygon, wherein: The minimum particle unit corresponding to the single photovoltaic panel or the detection output is recorded as a kth component, the viscera pollution proportion r k E [0,1] of the component is detected, and the component is attached with a non-negative weight w module for reflecting the importance of the component to the cleaning decision; Generation of array cluster boundaries Cj: Detecting all the component frames by using YOLOv, and writing the whole component frames into a binary mask M on the whole graph; performing single expansion on M by using a rectangular kernel K to connect adjacent/neighbor components; Extracting an outer contour of an expansion result, and obtaining an array cluster boundary Cj by using a polygon approximation method RDP; Each cluster contains a plurality of photovoltaic modules and weights thereof: wmodule＝Tier(rk)∈{1,2,3}; Step S3, path planning Step S3.1, constructing a solution, namely taking the heaviest cluster as a starting point, taking the current cluster as i, taking a candidate non-access cluster set as U, defining local cost for any j E U, namely setting cost (j) =λd (i, j) -Wj, namely the least cost, as the next point, sequencing the points, wherein d (i, j) is Euclidean distance between i and centroid of two array clusters of j, W j is array cluster weight,W 0 is the median of cluster weights, d 0 is the median of common neighboring cluster spacings; Step S3.2, local search, 2-opt and repositioning relocate are carried out on a single route, and the best improvement is carried out at the cost: The relocation relocate is used as a local search operation, any non-starting point node in the path is removed from the original position and inserted into other positions of the same path on the premise of not changing the starting point, the incremental cost delta is used as an evaluation criterion, and the best improvement principle is adopted to select whether to accept the movement; A movable object, which is any node v except a starting point; Allowed insertion position, all insertion slits in the route, but not allowing v to be inserted into the first position; The goal is to calculate the incremental Cost delta for each candidate insertion location using a defined Cost formula, delta = Cost (new) -Cost (old), take the optimal path, perform the move if delta <0, repeat until there is no improvement, Step S3.3, disturbance restarting, namely randomly moving out non-starting point nodes of p, and then reinserting one by one at the cost of the minimum increment so as to jump out of local search to find a better solution; S3.4, circulating until the number of the iterations is reached or the upper limit of no improvement is reached, and outputting an optimal cluster sequence; s3.5, merging the produced cluster sequences and removing the weight; step S