A method and system for large-scale inspection of photovoltaic power generation equipment

Abstract

The present invention discloses a large-scale inspection method and system for photovoltaic power generation equipment, which can improve the inspection efficiency of large-scale photovoltaic power stations, optimize the inspection route, and realize accurate positioning and identification of fault information. The method includes the following steps: monitoring and obtaining photovoltaic power generation equipment information in each area of ​​the photovoltaic power station, analyzing and giving inspection opinions; using an improved genetic algorithm to obtain the optimal inspection plan for drones according to the inspection opinions; inspecting photovoltaic power generation equipment based on the optimal inspection plan for drones; constructing a decentralized task assistance network based on the inspection situation to issue collaborative operation instructions to the inspected drones to realize large-scale inspection.

Description

Technical Field

[0001] The present invention belongs to the field of unmanned aerial vehicle path planning technology and fault detection technology, and specifically relates to a large-scale inspection method and system for photovoltaic power generation equipment. Background Art

[0002] At present, due to the huge consumption of traditional fossil fuel resources and energy depletion, facing the severe trend of energy crisis, we must be committed to vigorously developing and utilizing clean energy. To this end, a large number of photovoltaic power stations have been built to convert renewable light energy into electrical energy to meet the electricity needs of cities and rural areas.

[0003] With the access of photovoltaic power stations to the Internet of Things, remote monitoring of low power generation efficiency of a photovoltaic component, short circuit, short circuit, hot spot, etc. can automatically send commands to dispatch drones to designated locations for point-to-point inspections. However, as the scale of photovoltaic power station construction continues to increase, the efficiency of existing drone single-task inspection methods has decreased, and the optimal inspection path cannot be calculated. At the same time, the increase in the scale of photovoltaic power stations will also cause problems such as reduced inspection positioning accuracy and inability to accurately obtain fault information. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a large-scale inspection method and system for photovoltaic power generation equipment, which can improve the inspection efficiency of large-scale photovoltaic power stations, optimize the inspection routes, and realize accurate positioning and identification of fault information.

[0005] The present invention is achieved through the following technical solutions:

[0006] A large-scale inspection method for photovoltaic power generation equipment comprises the following steps:

[0007] Monitor and obtain information on photovoltaic power generation equipment in each area of ​​the photovoltaic power station, analyze and give inspection opinions;

[0008] According to the inspection opinions, an improved genetic algorithm is used to obtain the optimal inspection plan for UAVs;

[0009] Conduct inspections of photovoltaic power generation equipment based on the optimized inspection plan of drones;

[0010] Based on the inspection situation, a decentralized task assistance network is built to issue collaborative operation instructions to the inspection drones to achieve large-scale inspections.

[0011] Preferably, the photovoltaic power generation equipment information includes electrical parameters and natural condition parameters, wherein the electrical parameters include the current, voltage, power and load of the photovoltaic power generation equipment, and the natural condition parameters include the on-site temperature, humidity and wind parameters of each area.

[0012] Preferably, the monitoring acquires information of photovoltaic power generation equipment in each area of ​​the photovoltaic power station, analyzes and gives inspection opinions, specifically including:

[0013] The big data platform is used to collect statistics on the fault-prone areas and power generation component locations of photovoltaic power stations in the past to obtain historical data information;

[0014] The photovoltaic power generation equipment information is obtained by installing sensors in each area of ​​the photovoltaic power station;

[0015] Combine historical data information with the photovoltaic power generation equipment information obtained through measurement to provide inspection suggestions.

[0016] Preferably, the inspection opinion includes the fault location and fault condition of the photovoltaic power generation equipment in each area of ​​the photovoltaic power station.

[0017] Preferably, the optimization objective function expression of the drone optimization inspection scheme is:

[0018] Max(Score)=1 / (t×w t +s×w s )

[0019] Among them, Max(Score) is the optimization objective function, t represents the total inspection time, that is, the total time from the start of the inspection to the return of the last drone to the initial point; s represents the total inspection mileage, that is, the sum of the flight distances of all drones; w t and w s They are respectively expressed as the weight coefficients of the two indicators t and s.

[0020] Preferably, the method of obtaining the optimal inspection plan for the UAV by using an improved genetic algorithm according to the inspection opinions specifically includes:

[0021] Set the overall environmental state at the current time t to s t ,s t =(yn t ,loc t );

[0022] Among them, t ∈{0,1},yn t If it is 0, it means the area is not inspected. t If it is 1, it means the area has been inspected. t Indicates the current location information of the drone;

[0023] The drone’s current state data and the overall environment state s t As the input parameter of the improved genetic algorithm, the internal strategy π of the UAV itself t,jBased on the improved genetic algorithm, the neural network model is trained to obtain the corresponding output action u of each drone after optimization. t,j , multiple drone actions are combined to form a joint action u t ;

[0024] The overall environment state is based on the state transition function P(s t+1 ∣s t ,u t ) makes the corresponding environmental state transition and obtains the overall environmental state s at the next moment t+1 , repeat the above steps until the drone inspection mission is completed.

[0025] Preferably, the photovoltaic power generation equipment inspection based on the drone optimized inspection plan includes:

[0026] The UAV reaches the inspection location along the designated path according to the optimal inspection plan of the UAV, and uses template matching, image pyramid structure, Hough line detection or interference line filtering algorithm to extract the current status information of the photovoltaic power generation equipment, and uses the edge detection algorithm for fault diagnosis.

[0027] Preferably, the decentralized task assistance network constructed based on the inspection situation to issue collaborative operation instructions to the inspection drone specifically includes:

[0028] When a drone inspects the photovoltaic power generation equipment in the area and finds no faults, it uploads a task request to the decentralized task assistance network for release. The decentralized task assistance network issues assistance instructions to drones in the surrounding area. Drones in the surrounding area choose whether to accept the task based on their own status, and complete multiple drone collaborative inspections in sequence.

[0029] A large-scale inspection system for photovoltaic power generation equipment, comprising:

[0030] The information acquisition module is used to monitor and obtain information about photovoltaic power generation equipment in each area of ​​the photovoltaic power station, analyze it and give inspection opinions;

[0031] The inspection plan calculation module is used to obtain the optimal inspection plan for UAVs using an improved genetic algorithm based on inspection opinions;

[0032] UAV motion output module, used to inspect photovoltaic power generation equipment based on the optimal inspection plan of UAV;

[0033] The collaborative operation module is used to build a decentralized task assistance network based on the inspection situation to issue collaborative operation instructions to the inspection drones to achieve large-scale inspections.

[0034] Preferably, the UAV action output module includes an equipment inspection module, which is used for the UAV to reach the inspection position along the specified path according to the UAV optimized inspection plan, and uses template matching, image pyramid structure, Hough line detection or interference line filtering algorithm to extract the current state information of the photovoltaic power generation equipment, and uses the edge detection algorithm to perform fault diagnosis.

[0035] Compared with the prior art, the present invention has the following beneficial technical effects:

[0036] The present invention provides a large-scale inspection method for photovoltaic power generation equipment. The state of the photovoltaic power station is obtained through the Internet of Things monitoring system and corresponding inspection opinions are given. The unmanned aerial vehicle can conduct inspections with emphasis during inspections, thereby improving inspection efficiency. The unmanned aerial vehicle station dispatches the unmanned aerial vehicle to perform task path planning according to the unmanned aerial vehicle optimal inspection plan obtained by calculating the improved genetic algorithm, so that the inspection route is more reasonable and the unmanned aerial vehicle power is saved. In addition, the unmanned aerial vehicle used for inspection is equipped with a variety of detection modules, which can realize real-time inspection picture transmission. At the same time, the deep learning algorithm is used to identify hot spots, surface defects, etc., so as to accurately identify the fault location and determine the fault cause, and then make corresponding fault handling behaviors in time according to the fault location and fault cause, so as to avoid the failure of the photovoltaic power generation equipment affecting subsequent applications. While conducting inspections, a decentralized task assistance network was established based on the inspection situations fed back by the drones, and a task polling mechanism was introduced, so that each drone can work together at the same time during inspections, perform multiple tasks at the same time, conduct multi-angle inspections on difficult tasks, and obtain accurate fault information, saving task execution time as much as possible, ensuring the endurance of each drone and the completion speed of the overall inspection tasks to the greatest extent, so as to realize large-scale drone inspections and ultimately achieve the goal of reducing staff and increasing efficiency, increasing power generation, and increasing economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a front view of the drone in an embodiment of the present invention;

[0038] Figure 2 is a flow chart of a genetic algorithm in an embodiment of the present invention;

[0039] Figure 3 Schematic diagram of a multi-UAV inspection operation in an embodiment of the present invention;

[0040] Figure 4 It is a flow chart of the inspection method of the present invention.

[0041] In the picture, 1. UAV shell; 2. Communication antenna; 3. Wing part; 4. Visible light camera; 5. Thermal infrared imaging camera; 6. UAV bracket; 7. Photovoltaic panel assembly; 8. Inspection UAV. DETAILED DESCRIPTION

[0042] The principles and features of the present invention are further described in detail below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention. It should be noted that the drawings are all in a very simplified form and are not in precise proportions, and are only used to facilitate and clearly assist in explaining the purpose of the embodiments of the present invention.

[0043] It should be noted that when a component is referred to as being "fixed to" another component, it may be directly on the other component or there may be a central component. When a component is considered to be "connected to" another component, it may be directly connected to the other component or there may be a central component at the same time. When a component is considered to be "set on" another component, it may be directly set on the other component or there may be a central component at the same time.

[0044] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0045] The present invention provides a large-scale inspection method for photovoltaic power generation equipment, such as Figure 4 As shown, the following steps are included:

[0046] Monitor and obtain information of photovoltaic power generation equipment, analyze and give inspection opinions;

[0047] According to the inspection opinions, an improved genetic algorithm is used to obtain the optimal inspection plan for UAVs;

[0048] Conduct inspections of photovoltaic power generation equipment based on the optimized inspection plan of drones;

[0049] Based on the inspection situation, a decentralized task assistance network is built to issue collaborative operation instructions to the inspection drones to achieve large-scale inspections.

[0050] The present invention designs a large-scale inspection method for photovoltaic power generation equipment. The Internet of Things monitoring system obtains the status of the photovoltaic power station and gives corresponding inspection opinions. The drone can conduct inspections with emphasis during inspections, thereby improving inspection efficiency. The drone station dispatches the drone to plan the task path according to the drone optimization inspection plan obtained by calculating the improved genetic algorithm, making the inspection route more reasonable and saving drone power. The drone used for inspection is equipped with a variety of detection modules, which can realize real-time inspection picture transmission. At the same time, the deep learning algorithm is used to identify hot spots, surface defects, etc., so as to accurately identify the fault location and determine the fault cause, and then make corresponding fault handling behaviors in time according to the fault location and fault cause, so as to avoid the failure of photovoltaic power generation equipment affecting subsequent applications. While conducting inspections, a decentralized task assistance network was established based on the inspection situations fed back by the drones, and a task polling mechanism was introduced, so that each drone can work together at the same time during inspections, perform multiple tasks at the same time, conduct multi-angle inspections on difficult tasks, and obtain accurate fault information, saving task execution time as much as possible, ensuring the endurance of each drone and the completion speed of the overall inspection tasks to the greatest extent, so as to realize large-scale drone inspections and ultimately achieve the goal of reducing staff and increasing efficiency, increasing power generation, and increasing economic benefits.

[0051] The large-scale inspection method for photovoltaic power generation equipment described in the present invention has major improvements including remote Internet of Things monitoring system monitoring equipment information, path planning algorithm design and task scheduling network construction.

[0052] The IoT monitoring system obtains the status of power generation components in each area of ​​the photovoltaic power station by setting up a large number of sensors in the plant area, including electrical parameters such as current, voltage, power, and load, as well as natural parameters such as on-site temperature and humidity parameters in each area and wind parameters. The big data platform summarizes the areas where faults frequently occurred in the past and the locations of power generation components, and gives inspection opinions, namely the fault location, fault conditions, and other information in combination with the current electrical parameters and natural conditions.

[0053] The drone station receives inspection opinions from the IoT monitoring system, and based on the inspection opinions, uses an improved genetic algorithm to obtain the optimal inspection plan. According to the optimal plan, the drone is dispatched for inspection.

[0054] Furthermore, drone stations are set up around the photovoltaic power station. According to the specific range of the photovoltaic power station and the maximum inspection range of the drone, the photovoltaic power station is divided to obtain the appropriate number of drone stations and configure the appropriate number of drones.

[0055] Furthermore, the UAV adopts a multi-rotor UAV, which has good stability and can achieve vertical take-off and landing.

[0056] Furthermore, the drone is equipped with photoelectric sensors, which can achieve fixed-altitude cruising and facilitate the photography of photovoltaic panels. It is equipped with a high-definition photoelectric camera, which can perform automatic and manual zoom. It is equipped with a thermal infrared camera, which can photograph the hot spots on the photovoltaic panels.

[0057] Furthermore, the drone is equipped with GPS and can reach the designated PV module location. The drone exchanges information with surrounding drones and drone stations through ZigBee wireless communication.

[0058] Furthermore, the drone extracts photovoltaic components through algorithms such as template matching, image pyramid structure, Hough line detection, and interference line filtering. Fault diagnosis is performed using algorithms such as edge detection. Environmental interference is filtered out through near-circularity, component grayscale variance, etc., and parameter adaptation is attempted according to the shooting environment.

[0059] Furthermore, when the drone runs out of battery and cannot complete the inspection mission, it can return to the nearest drone station for charging. The drone station can dispatch a fully charged drone to take over the unfinished mission and retain the previous mission information.

[0060] The optimization of the UAV inspection route is solved by using an improved genetic algorithm:

[0061] In order to measure the pros and cons of the inspection method scheme, the following indicators are used:

[0062] ① Total inspection time t: the time from the start of the inspection to the last drone returning to the initial point;

[0063] ②Total inspection mileage s: the sum of the flight distances of all drones.

[0064] The weight coefficients of the two indicators, total inspection time t and total inspection mileage s, are w t and w s , the total score of the scheme is expressed by the following formula: Score = 1 / (t×w t +s×w s ). The larger the score, the better, which is the optimization objective function of the solution.

[0065] In specific inspections, the task processing process of the drone may change during the inspection task and needs to be adjusted at any time. The photovoltaic power generation area is divided into regions, and the drone station dispatches a corresponding number of drones for inspection according to the number of fault points in different regions. The entire task environment can be abstracted as a partially observable Markov decision process, and the process of drones flying to photovoltaic power generation equipment and inspecting is regarded as an abstract action. On this basis, the task is analyzed and planned, and the multi-drone task planning problem is solved by improving the genetic algorithm learning algorithm through multiple drones.

[0066] Assume the overall environmental state is st ,s t =(yn t ,loc t ), where yn t =(yn t,1 ,yn t,2 ,…yn t,N ), the arrangement order is the optimal solution obtained by the aforementioned genetic algorithm, yn t,i ∈{0,1}, 0 represents the uninspected part, and 1 represents the inspected part. t Indicates the location information of multiple drones, i.e. loc t =(loc t,3 ,loc t,2 ,…loc t,M ). Each drone’s own state observation is O t,j , o t,j =dyn t,j ,loc t,j ), where dyn t,j Indicates the inspection status of all facilities within the drone's range X at time t, loc t,j is the current position state of each drone. According to its internal strategy π t,j , get the output action u t,j , multiple drone actions are combined to form a joint action u t , the environment will be based on the state transition function P(s t+1 ∣s t ,u t ) Make the corresponding environmental state transition and get the next state s t+1 , and repeat this cycle until the mission is completed. t ,u t ,r t ,s t+1 >, spatial position information and other data are used as input parameters of the improved genetic algorithm to train the neural network model, obtain the corresponding action output of the drone, and accumulate experience for the next inspection.

[0067] Furthermore, each drone can see multiple agents, obtain its own status from the current environment, and perform output actions according to its own internal strategy.

[0068] Furthermore, multiple drones form joint actions through combined actions, and the environment will be converted into a function according to the state, making corresponding environmental state transfers to obtain the state at the next moment.

[0069] Furthermore, all photovoltaic power generation equipment to be inspected may be arranged in task order, and a {0, 1} sequence may be used to distinguish whether the inspection task is completed. ​

[0070] Furthermore, each drone can obtain the inspection status of surrounding photovoltaic power generation equipment and plan the next task.

[0071] Furthermore, the information from drone inspections is stored and sent to the cloud server, so that the staff at the main control center can further diagnose the cause of the photovoltaic equipment failure.

[0072] Furthermore, the UAV flight control part is loaded with an improved genetic algorithm, and a deep learning network is used to represent the value function to rationally plan multiple UAV tasks.

[0073] Furthermore, the drone has automated modeling, autonomous flight path planning, one-click autonomous inspection, is not restricted by terrain and environment, supports power-off continued flight, edge computing, fault diagnosis, automatic identification of fault types, fault point information management, automatic generation of inspection reports, big data operation and maintenance knowledge base, and guidance of on-site operation and maintenance.

[0074] At the same time, a task polling mechanism is introduced into the inspection method described in the present invention. When a drone inspects a photovoltaic panel and no fault result is found, a task request can be uploaded to the task center for publication. The task center issues assistance instructions to the surrounding drones. The surrounding drones choose whether to accept the task based on their own status, and the coordinated inspection of multiple drones is completed in sequence.

[0075] Furthermore, each drone can perform collaborative work during inspections. For example, two or more drones on the same mission path can perform multi-task processing on the same photovoltaic module at the same time. For example, when one drone is transmitting images on site, another drone can perform infrared hot spot detection, and the third drone can perform visual inspections of surface structural damage. Performing multiple tasks at the same time can maximize the endurance of each drone and the speed at which the overall inspection task is completed.

[0076] Example

[0077] In an embodiment of the present invention, a method and system for large-scale inspection of photovoltaic power generation equipment are proposed. The inspection system includes a drone station and multiple drones therein, an Internet of Things monitoring system, and a decentralized task assistance network.

[0078] Reference Figure 1 The drone in this embodiment adopts a quad-rotor design, and each drone is equipped with a communication device that can communicate with nearby drones. The drone is a low-cost operation and maintenance equipment that can quickly reach the inspection location based on the GPS signal, and uses algorithms such as template matching, image pyramid structure, Hough line detection, and interference line filtering to extract photovoltaic module information, and then uses edge detection and other algorithms for fault diagnosis.

[0079] The drone in this embodiment is also equipped with a suitable visual sensor, such as a photoelectric sensor, which can realize the drone's fixed-altitude flight and hovering detection; and is equipped with a thermal infrared imaging camera 5 and a visible light imaging camera 4, with a real-time image transmission distance of up to 15 km. It also uses image processing technology and photovoltaic component fault detection technology to automatically detect abnormal conditions such as component dust, dirt, cracks, occlusions, hot spots, etc., and report abnormal details and precise location information.

[0080] In this embodiment, each drone has its own task arrangement and obeys the task dispatch of the drone station. After completing the task, it returns to the nearest drone station. If the task cannot be completed during the mission execution and the battery is insufficient, the drone station will dispatch a new drone to take over the task and retain the task data.

[0081] Reference Figure 2 ,In this embodiment, the initialization task information includes the location of the fault point, the ,fault type prediction information, and the location of photovoltaic panels that may have hidden ,hazards are integrated to establish an n+m-1 dimensional array, where n is the number of all ,points to be detected, and m is the number of photovoltaic ,field areas.

[0082] The first n+m-1 elements are randomly arranged, and a specific number of cases are extracted based on experience to obtain the first generation of the population. The score of each case in the first generation is calculated, and the next generation of the population is obtained through the selection, crossover, and mutation operations in the genetic algorithm.

[0083] When performing the selection operation, all individuals in the initialized scheme are first sorted from small to large according to the optimization objective function Score, and then a certain proportion of individuals are selected to be directly retained in the next generation.

[0084] After that, two different individuals are randomly selected from the remaining individuals, their scores are compared, and the individual with the smaller score is directly retained to the next generation. After executing the above two steps, the number of individuals in the population is reduced by half, and the population size will be maintained unchanged through crossover and mutation operations.

[0085] Among them, the mutation operation is an operation performed on a single chromosome to generate a new individual. When performing the mutation operation, the chromosome to be mutated is first randomly determined, and then two different positions in the chromosome are randomly selected and the values ​​of the elements in these two positions are exchanged. The mutation operation may occur in the same section, which can change the inspection order of the mission points of a drone; it may also occur in different sections, which can change the distribution of mission points between drones of different sorties, thereby obtaining the optimal solution.

[0086] In this embodiment, the number of drone stations and drones is set accordingly according to the scale of the photovoltaic power station. Usually, the drone station is set in a flat area around the power station to facilitate the staff to repair and inspect the drones; the drone station can be supplied with electricity through the photovoltaic power station; the drone station can communicate with the remote Internet of Things system through the nearby 5G base station to report the operation status of the drone station and the status of the drone equipment.

[0087] Reference Figure 3 In this embodiment, there are 10 fault points in the current inspection task. First, the task information is initialized, including the location information of the fault point and the possible fault conditions. The inspection site is divided into three areas. The genetic algorithm is used to plan the inspection route of the drone. A 12 (10 + 3 - 1) bit array is established as the task plan. The initial optimal plan is [1, 4, 8, 5, 9, 2, 0, 3, 7, 0, 6, 10]. This array indicates that the inspection order of the fault points in the first area is [1, 4, 8, 5, 9, 2], the inspection order in the second area is [3, 7], and the inspection order in the third area is [6, 10]. At this time, the formula is used to calculate Score = 1 / 100 (min) × 0.4 + 19.2 (km) × 0.6 = 0.0281.

[0088] During the actual inspection, due to the different task processing situations at different fault points, the UAV autonomously performs secondary path planning and obtains the actual optimal solution of [1,4,8,5,2,0,3,7,0,6,9,10]. At this time, the score Score (best) = 1 / 100 (min) × 0.4 + 14 (km) × 0.6 = 0.0308.

[0089] When a drone fails or runs out of power in a certain inspection area, it can upload mission completion information to the mission assistance network and issue assistance instructions. Nearby drones can choose whether to respond to the mission assistance request based on their own work conditions. After accepting the mission request, they can reach the mission request point to continue the unfinished inspection work.

[0090] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any ordinary technician in the industry can smoothly implement the present invention as shown in the drawings and described above. However, any equivalent changes, modifications and evolutions made by technicians familiar with the profession without departing from the scope of the technical solution of the present invention using the technical content disclosed above are all equivalent embodiments of the present invention. At the same time, any equivalent changes, modifications and evolutions made to the above embodiments based on the essential technology of the present invention are still within the protection scope of the technical solution of the present invention.

Claims

1. A large-scale inspection method for photovoltaic power generation equipment, characterized in that: The steps include: Monitor and obtain information on photovoltaic power generation equipment in each area of ​​the photovoltaic power station, analyze and give inspection opinions; According to the inspection opinions, the improved genetic algorithm is used to obtain the optimal inspection plan of the UAV; the optimization objective function expression of the optimal inspection plan of the UAV is: Max(Score)=1 / (t× +s× ) Among them, Max (Score) is the optimization objective function, t represents the total inspection time, that is, the total time from the start of the inspection to the return of the last drone to the initial point; s represents the total inspection mileage, that is, the sum of the flight distances of all drones; and They are respectively expressed as the weight coefficients of the two indicators t and s; Conduct inspections of photovoltaic power generation equipment based on the optimized inspection plan of drones; Based on the inspection situation, a decentralized task assistance network is built to issue collaborative operation instructions to the inspection drones to achieve large-scale inspections, including: When a drone inspects the photovoltaic power generation equipment in the area and finds no faults, it uploads a task request to the decentralized task assistance network for release. The decentralized task assistance network issues assistance instructions to drones in the surrounding area. Drones in the surrounding area choose whether to accept the task based on their own status, and complete multiple drone collaborative inspections in sequence.

2. A large-scale inspection method for photovoltaic power generation equipment according to claim 1, characterized in that: The photovoltaic power generation equipment information includes electrical parameters and natural condition parameters, wherein the electrical parameters include the current, voltage, power and load of the photovoltaic power generation equipment, and the natural condition parameters include the on-site temperature, humidity and wind parameters of each area.

3. A large-scale inspection method for photovoltaic power generation equipment according to claim 1, characterized in that: The monitoring obtains information about photovoltaic power generation equipment in each area of ​​the photovoltaic power station, analyzes and gives inspection opinions, specifically including: The big data platform is used to collect statistics on the fault-prone areas and power generation component locations of photovoltaic power stations in the past to obtain historical data information; The information of photovoltaic power generation equipment is obtained by installing sensors in each area of ​​the photovoltaic power station; Combine historical data information with the photovoltaic power generation equipment information obtained through measurement to provide inspection suggestions.

4. A large-scale inspection method for photovoltaic power generation equipment according to claim 1, characterized in that: The inspection opinions include the fault locations and fault conditions of the photovoltaic power generation equipment in each area of ​​the photovoltaic power station.

5. A large-scale inspection method for photovoltaic power generation equipment according to claim 1, characterized in that: The method of using the improved genetic algorithm to obtain the optimal inspection plan for the UAV according to the inspection opinions specifically includes: Assume the overall environmental state at the current time t is , ; in, {0,1}, If it is 0, it means the area is not inspected. If it is 1, it means the area has been inspected. Indicates the current location information of the drone; The drone's own state data and overall environment state at the current time t As the input parameter of the improved genetic algorithm, the internal strategy of the UAV itself The neural network model is trained based on the improved genetic algorithm to obtain the corresponding output action of each drone after optimization. , multiple drone actions are combined to form a joint action ; The overall environment state is based on the state transition function Make the corresponding environmental state transition to obtain the overall environmental state at the next moment , repeat the above steps until the drone inspection mission is completed.

6. A large-scale inspection method for photovoltaic power generation equipment according to claim 1, characterized in that: The photovoltaic power generation equipment inspection based on the drone optimization inspection plan includes: The UAV reaches the inspection location along the designated path according to the optimal inspection plan of the UAV, and uses template matching, image pyramid structure, Hough line detection or interference line filtering algorithm to extract the current status information of the photovoltaic power generation equipment, and uses the edge detection algorithm for fault diagnosis.

7. A large-scale inspection system for photovoltaic power generation equipment, characterized in that: The inspection method for implementing any one of claims 1 to 6 comprises: The information acquisition module is used to monitor and obtain information about photovoltaic power generation equipment in each area of ​​the photovoltaic power station, analyze it and give inspection opinions; The inspection plan calculation module is used to obtain the optimal inspection plan for UAVs using an improved genetic algorithm based on inspection opinions; UAV motion output module, used to inspect photovoltaic power generation equipment based on the optimal inspection plan of UAV; The collaborative operation module is used to build a decentralized task assistance network based on the inspection situation to issue collaborative operation instructions to the inspection drones to achieve large-scale inspections.

8. A large-scale inspection system for photovoltaic power generation equipment according to claim 7, characterized in that: The UAV action output module includes an equipment inspection module, which is used for the UAV to reach the inspection position along the specified path according to the UAV optimization inspection plan, and uses template matching, image pyramid structure, Hough line detection or interference line filtering algorithm to extract the current state information of the photovoltaic power generation equipment, and uses the edge detection algorithm to perform fault diagnosis.

Images