CN-121994225-A - Photovoltaic scene-oriented weighted time-delay unmanned aerial vehicle cleaning optimal path judging method

CN121994225ACN 121994225 ACN121994225 ACN 121994225ACN-121994225-A

Abstract

The invention relates to a method for judging a cleaning optimal path of a weighted time-delay unmanned aerial vehicle oriented to a photovoltaic scene. Taking the component-level weighted delay WLD as a main target, and introducing indexes such as early benefit of the first 50% of time and the like to assist assessment for preventing parallel situations. The method can truly reflect the benefit value of 'earlier completed more important objects', can still stably sort and select the candidate paths when a large number of candidate paths exist, automatically bias the paths covering more high-weight components in a short time when component-level weighting delay WLD are parallel, remarkably improve early power generation recovery benefits, can be directly embedded into the existing system, is also convenient for subsequent replacement or superposition of stronger path generators, simultaneously gives out the result representation of segment levels and point levels of the scored optimal paths, provides a deriving flow, and is convenient for an unmanned aerial vehicle upper computer to directly execute and subsequently review.

Inventors

WANG DASHUAI
XU XIWEN
Liu Chuanghai
FENG XIAOYANG
Geng changxing

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. A method for judging a cleaning optimal path of a weighted delay unmanned aerial vehicle oriented to a photovoltaic scene 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, optimizing the generated candidate cleaning paths Each component is provided with corresponding weight And the cumulative time to complete the current component Ensuring the dual constraint of dirt priority and flight time, wherein c is the c-th array cluster, p is the p-th component in the array cluster, and the specific grading rule is used for carrying out step S4; step S4, candidate global path scoring master objective function weighting delay WLD Step S4.1, calculating cluster/component completion time for each candidate global path, defining a weighted delay WLD target, and scoring the candidate paths: component level weighted delay WLD: step S4.2 Multi-index scoring and screening of candidate Global paths While keeping the weighted delay WLD as the main target, to encourage early coverage of high-weight components, improve the efficiency of the action, the following indexes are set and used together with WLD for evaluation and ranking of candidate paths: Early benefit e_module50%: sum of component weights cleaned in the first 50% of time; Cleaning duty cycle Wherein, T clean is the accumulated cleaning time after cleaning the current component, and T total is the accumulated time before cleaning the current component; Step S4.3 candidate Global Path scoring procedure 1) Firstly, minimizing WLD module ; 2) If parallel, maximizing early benefit E_Module50%; 3) Comparing the cleaning duty ratio P clean ; step S5, path achievement expression and derivation Segment expression, namely outputting two types of line segments according to actual execution actions: A MOVE section, namely a last thick red line exit-a next thin gray line entry; Clear section, fine gray line entry→thick red line exit; pixel coordinates and longitude and latitude are derived by projecting the affine transformation+crs to WGS84 (EPSG: 4326) if the source data is GeoTIFF including CRS.
2. The method for determining the optimal path for cleaning the weighted delay unmanned aerial vehicle for the photovoltaic scene 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. The method for determining the optimal path for cleaning the weighted delay unmanned aerial vehicle for a photovoltaic scene according to claim 1, wherein in the step S5, when the pixel coordinates and the longitude and latitude are derived: segment-level CSV with start and end points for each segment in the cleaning route. Point-level CSV, namely, the endpoints of each segment are orderly arranged; And (3) visualization, namely superposing the moving path and the cleaning path on the base map, and numbering.

Description

Photovoltaic scene-oriented weighted time-delay unmanned aerial vehicle cleaning optimal path judging method Technical Field The invention relates to the technical field of unmanned aerial vehicle operation and maintenance and path optimization, in particular to a cleaning unmanned aerial vehicle cleaning optimal path optimization and judgment method for field operation of a photovoltaic power station. Background The existing photovoltaic unmanned aerial vehicle cleaning mostly uses the shortest flight path as a constraint condition, for a photovoltaic power station, the pollution severity of different components/arrays is different, and the higher the pollution degree is, the higher the cleaning priority of the photovoltaic components is, so that the optimal power generation recovery benefit cannot be brought only by taking the shortest path as the constraint. Therefore, a unified evaluation index weighting Delay (WLD, weighted Latency of Delay) capable of simultaneously measuring the dirt weight and the completion time is needed, and meanwhile, multiple indexes such as early benefits are introduced, so that the key targets are ensured to be covered preferentially under the same WLD, the optimal cleaning paths are screened out from the candidate paths, and the cleaning benefits and the working efficiency are improved. Disclosure of Invention The invention aims to overcome the problems in the prior art and provides a method for judging the cleaning optimal path of a weighted delay unmanned aerial vehicle oriented to a photovoltaic scene, which is used for uniformly and comparably quantitatively scoring and sequencing candidate global paths of any source on the premise of not limiting a specific generation algorithm. The method comprises the steps of taking a component-level weighted delay WLD as a main target, introducing indexes such as early benefits of the first 50% of time and the like for auxiliary evaluation to prevent paralleling, and achieving that 1, the component-level weighted delay WLD is taken as a global main target, grading and ranking candidate paths, 2, when the component-level weighted delay WLD is in paralleling, paths with higher early benefits (such as component weights and larger after cleaning in the first 50% of time) are preferentially selected, 3, auxiliary indexes such as cleaning duty ratio and the like are supported to be added to form a stable judging flow under multiple indexes, and 4, finally, outputting an optimal path, and deriving pixel coordinates and longitude and latitude of a traversing point for cleaning use of a follow-up unmanned aerial vehicle. In order to achieve the technical purpose and the technical effect, the invention is realized by the following technical scheme: A method for judging a cleaning optimal path of a weighted delay unmanned aerial vehicle oriented to a photovoltaic scene 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, optimizing the generated candidate cleaning paths Each component is provided with corresponding weightAnd the cumulative time to complete the current componentEnsuring the dual constraint of dirt priority and flight time, wherein c is the c-th array cluster, p is the p-th component in the array cluster, and the specific grading rule is used for carrying out step S4; step S4, candidate global path scoring master objective function weighting delay WLD Step S4.1, calculating cluster/component completion time for each candidate global path, defining a weighted delay WLD target, and scoring the candidate paths: component level weighted delay WLD: step S4.2 Multi-index scoring and screening of candidate Global paths While keeping the weighted delay WLD as the main target, to encourage early coverage of high-weight components, im