A method, system, medium and program product for drone inspection of photovoltaic stations

Abstract

A method, system, medium, and program product for drone inspection of photovoltaic stations, relating to the general field of control or regulation systems, includes: controlling a light drone to conduct inspections along a preset inspection route, acquiring real-time inspection images and actual flight parameters; determining the real-time inspection progress and fault detection areas based on the real-time inspection images, and calculating a predicted inspection duration; determining multiple target areas to be inspected, including historical fault areas and fault detection areas, and constructing a real-time re-inspection route; calculating equivalent flight parameters of a standard drone based on the actual flight parameters of the light drone during the inspection, and calculating a predicted re-inspection duration for the standard drone to complete the inspection along the real-time re-inspection route; and when the predicted re-inspection duration is greater than or equal to the predicted inspection duration, controlling the standard drone to conduct inspections along the real-time re-inspection route, acquiring inspection and re-inspection images, and generating an inspection report. Implementing this application can improve the efficiency of drone inspections at photovoltaic stations.

Description

Technical Field

[0001] The present application relates to the general field of control or regulation systems, and in particular to a method, system, medium and program product for drone inspection of photovoltaic stations. Background Art

[0002] With the rapid development of photovoltaic power generation technology, the construction of large-scale photovoltaic power plants continues to expand. To ensure the stable operation and power generation efficiency of photovoltaic power plants, regular inspections are required to promptly detect and address problems such as component failures, hot spots, and stains. Traditional manual inspections are inefficient and pose safety risks, leading to the widespread use of drone inspection technology in photovoltaic power plants.

[0003] In related technologies, drone inspections of photovoltaic plants primarily use drones equipped with infrared cameras, visible light cameras, and other detection equipment to inspect photovoltaic arrays along pre-set routes. The drones fly along the designated routes, collecting image data using onboard equipment and transmitting this data to ground stations for fault analysis and resolution.

[0004] However, during the inspection process, especially in scenarios where inspection requirements are becoming increasingly sophisticated, drones need to carry a variety of inspection equipment, which results in a heavy weight and limited flight time. To ensure inspection accuracy, drones need to fly slowly and steadily at a relatively low altitude, further increasing power consumption. Therefore, when faced with large-scale photovoltaic inspections, drones in related technologies often require multiple battery replacements and recharging, resulting in longer overall inspection times and lower efficiency. Summary of the Invention

[0005] The present application provides a method, system, medium and program product for drone inspection of photovoltaic stations, which are used to improve the inspection efficiency of drones in photovoltaic stations.

[0006] In the first aspect, the present application provides a method for drone inspection of photovoltaic stations, which is applied to the station inspection system, wherein the station inspection system includes a light drone and a standard drone, wherein the number of onboard equipment of the light drone is less than that of the standard drone, and the weight of the light drone is lower than that of the standard drone; the method comprises: controlling the light drone to inspect according to a preset inspection route, obtaining real-time inspection images and actual flight parameters; the actual flight parameters include the actual flight speed and heading adjustment data of each area segment; determining the real-time inspection progress and the fault detection area according to the real-time inspection image, and determining the predicted inspection time after the light drone inspection is completed based on the real-time inspection progress; inspection time; determine multiple target areas to be inspected including historical fault areas and fault detection areas, and construct a real-time re-inspection route including multiple target areas to be inspected; calculate the equivalent flight parameters of the standard UAV inspection under the same environmental conditions based on the actual flight parameters of the light UAV during the inspection; calculate the predicted re-inspection time for the standard UAV inspection to be completed on the real-time re-inspection route based on the regional positions and equivalent flight parameters of multiple target areas to be inspected; when the predicted re-inspection time is greater than or equal to the predicted inspection time, control the standard UAV to inspect according to the real-time re-inspection route to obtain an inspection review image; generate an inspection report based on the real-time inspection image and the inspection review image.

[0007] In the above embodiment, the station inspection system uses light drones to conduct initial inspections and obtain real-time parameters, and combines this with the re-inspection mechanism of standard drones to achieve optimal allocation of inspection resources. Light drones can quickly complete initial inspections of large areas due to their lightweight characteristics, while standard drones focus on precise inspections of key areas. The system analyzes the inspection progress and fault areas in real time, dynamically plans re-inspection routes, and ensures the synergy of the two types of drones. This hierarchical detection strategy not only ensures inspection efficiency but also maintains detection accuracy, significantly improving the inspection efficiency of photovoltaic stations.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the step of calculating the equivalent flight parameters of a standard UAV during inspection under the same environmental conditions based on the actual flight parameters of the light UAV during inspection specifically includes: based on the power parameter configuration of the light UAV and the standard UAV, constructing a flight operation mapping of the two models when responding to the same wind condition environment; based on the inspection parameter configuration of the light UAV and the standard UAV, constructing a detection time mapping of the two models when processing the same inspection object; based on the flight operation mapping and the detection time mapping, converting the actual flight parameters of the light UAV into equivalent flight parameters of the standard UAV.

[0009] In the above embodiment, the station inspection system realizes the precise conversion of flight parameters between different UAVs by establishing a flight operation mapping and detection time mapping relationship between the two types of aircraft; the system analyzes the flight characteristics under the same wind conditions based on the power parameter configuration, and considers the inspection parameter configuration to evaluate the detection time difference, thereby accurately predicting the actual operation performance of the standard UAV, reducing the difficulty of real-time processing of control parameters when the standard UAV is working, and improving the execution efficiency of the inspection task.

[0010] In combination with some embodiments of the first aspect, in some embodiments, the step of calculating the predicted re-inspection time for completing the inspection of a standard UAV on a real-time re-inspection route based on the regional positions and equivalent flight parameters of multiple target areas to be inspected specifically includes: dividing the real-time re-inspection route into an inspection section and a transition section based on the regional positions of multiple target areas to be inspected; the inspection section is located in the target area to be inspected, and the transition section is located between adjacent target areas to be inspected; calculating the inspection section time of the standard UAV in the inspection section based on the equivalent flight parameters; calculating the transition section time of the standard UAV in the transition section based on the equivalent flight parameters and the power constraints of the standard UAV; accumulating the inspection section time and the transition section time to obtain the predicted re-inspection time for completing the inspection of the standard UAV.

[0011] In the above embodiment, the site inspection system subdivides the re-inspection route into inspection sections and transition sections, and combines equivalent flight parameters to perform accurate time prediction; the system calculates the time taken by the standard drone in each section separately, and obtains the overall predicted duration by accumulation, thereby improving the accuracy of time prediction.

[0012] In combination with some embodiments of the first aspect, in some embodiments, the step of calculating the transition section time of the standard UAV in the transition section based on the equivalent flight parameters and the power constraints of the standard UAV specifically includes: calculating the remaining power of the standard UAV after the last inspection section based on the battery capacity and power consumption characteristics of the standard UAV; when the remaining power is lower than the preset power threshold, determining the battery swap position so that the standard UAV performs the battery swap operation before the next inspection section; determining the maximum flight speed of the standard UAV to and from the battery swap position based on the equivalent flight parameters and the power constraints of the standard UAV, and calculating the battery swap path time; calculating the transition section time of the standard UAV based on the battery swap path time, the equipment start and stop time, and the expected battery swap time.

[0013] In the above embodiment, the station inspection system monitors the power status of the standard drone in real time and intelligently plans the battery replacement operation when the remaining power is insufficient; the system comprehensively considers factors such as the battery replacement path time and the equipment start and stop time to ensure that the standard drone can operate continuously and stably, effectively avoiding mission interruption due to insufficient power.

[0014] In combination with some embodiments of the first aspect, in some embodiments, before the step of controlling the standard UAV to perform inspections according to the real-time re-inspection route and obtaining the inspection review image, the method also includes: calculating the re-inspection time nodes when the standard UAV arrives at multiple target inspection areas when performing inspections according to the real-time re-inspection route; calculating the inspection time nodes when the light UAV arrives at multiple target inspection areas based on the real-time inspection progress; when the re-inspection time node is earlier than the inspection time node, marking the corresponding target inspection area as the crew overlap area; determining the re-inspection take-off delay time so that the re-inspection time node of the crew overlap area is later than the inspection time node.

[0015] In the above embodiment, the station inspection system avoids resource waste caused by overlapping units by calculating and comparing the working time nodes of the two types of drones; the system adjusts the re-inspection take-off time to ensure that the standard drone performs re-inspection after the light drone completes the initial inspection, thereby achieving efficient coordination of the two types of drones.

[0016] In combination with some embodiments of the first aspect, in some embodiments, when the re-inspection time node is earlier than the inspection time node, after the step of marking the corresponding target area to be inspected as the unit overlap area, the method also includes: determining one or more new areas to be inspected based on the real-time inspection image on the preceding route corresponding to the unit overlap area on the real-time re-inspection route; recalculating the re-inspection time node of the unit overlap area based on the newly added areas to be inspected, so that the re-inspection time node of the unit overlap area is later than the inspection time node.

[0017] In the above embodiment, the station inspection system dynamically adjusts the re-inspection route, fully utilizes the waiting time of the standard drone, and adds new inspection tasks to the front route in the unit overlap area, which not only avoids unit overlap but also improves equipment utilization.

[0018] In combination with some embodiments of the first aspect, in some embodiments, after the step of generating an inspection report based on real-time inspection images and inspection review images, the method also includes: extracting fault image features from the inspection review images to form a fault list; matching the fault types of multiple detected faults in the fault list with the power generation impact parameter table preset by the photovoltaic station to obtain the fault impact degree values ​​of the detected faults; sorting the detected faults according to the fault impact degree values ​​to obtain a task list of maintenance tasks corresponding to the detected faults; generating fault location coordinates and maintenance instructions corresponding to multiple maintenance tasks in the task list.

[0019] In the above embodiment, the station inspection system systematically analyzes and processes the inspection results, matches the detected faults with parameters affecting power generation, generates a prioritized maintenance task list, and provides accurate fault locations and maintenance guidance, thereby improving the efficiency of subsequent maintenance work.

[0020] In a second aspect, an embodiment of the present application provides a site inspection system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the site inspection system to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer program product comprising instructions. When the computer program product is run on a site inspection system, the site inspection system executes the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium comprising instructions. When the instructions are executed on a site inspection system, the site inspection system executes the method described in the first aspect and any possible implementation of the first aspect.

[0023] It is understood that the station inspection system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects that can be achieved can be referenced to the beneficial effects of the corresponding methods and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0025] 1. The dual-drone inspection solution uses a light drone for rapid initial inspections and works in conjunction with a standard drone. This allows the two drones to fully leverage their respective strengths for complementary collaboration, effectively resolving the existing issue of a single drone being unable to balance inspection efficiency and detection accuracy. This significantly improves the efficiency and quality of PV station inspections. Light drones are equipped with basic detection equipment, are lightweight, and have long flight times, enabling them to quickly complete initial inspections of large-scale sites. Standard drones are equipped with high-precision detection equipment and focus on precise re-inspections of key areas. Real-time analysis of initial inspection data and dynamic planning of re-inspection routes ensure the efficient and coordinated operation of the two drones, avoiding waste of equipment resources while ensuring inspection quality.

[0026] 2. Due to the adoption of a scheme that constructs a mapping relationship between the two aircraft models based on the power parameter configuration and the inspection parameter configuration, the flight performance of the standard UAV in actual work can be accurately predicted, which effectively solves the problem in the existing technology that it is impossible to accurately evaluate the differences in the operating capabilities of different aircraft models, thereby realizing the precise planning and efficient execution of the re-inspection task; by analyzing the differences in the flight characteristics of the two UAVs under the same wind conditions, a flight operation mapping relationship is established; at the same time, based on the performance differences between the two models when handling the same inspection objects, a detection time mapping relationship is constructed; the actual flight data of the light UAV can be accurately converted into equivalent parameters of the standard UAV, providing a reliable basis for re-inspection planning.

[0027] 3. Due to the adoption of a scheme that divides the re-inspection route into inspection sections and transition sections for segmented calculation, accurate time prediction can be made based on the characteristics of different sections, effectively solving the problem of inaccurate re-inspection duration prediction in existing technologies leading to inefficient task planning, thereby achieving optimized execution of inspection tasks; first, the route is divided into different sections according to the distribution of the target area to be inspected, and then the operating time of the standard UAV in each section is calculated separately; in the inspection section, the speed limit brought about by the detection accuracy requirements is mainly considered, and in the transition section, the impact of power constraints on flight speed is mainly considered to ensure work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a flow chart of a method for drone inspection of a photovoltaic station in an embodiment of the present application;

[0029] Figure 2 This is another flowchart of the drone inspection method for photovoltaic stations in an embodiment of the present application;

[0030] Figure 3 It is a schematic diagram of the physical device structure of the station inspection system in the embodiment of the present application. DETAILED DESCRIPTION

[0031] The terms used in the following examples of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application, the singular expressions "a", "an", "above", "the", and "this" are intended to include plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations of one or more of the listed items.

[0032] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more.

[0033] For ease of understanding, the application scenarios of the embodiments of the present application are introduced below.

[0034] A large-scale photovoltaic power station covers an area of ​​2,000 mu (approximately 1,000 acres) and contains over 500,000 photovoltaic modules. Traditional manual inspections only complete 20% of the area each month, resulting in numerous failures going undetected. Especially during the high temperatures of summer, modules are prone to failures such as hot spots and cracks. If not detected and addressed promptly, these failures can lead to a rapid decline in module efficiency and even pose a fire hazard. The power station's operations and maintenance department urgently needed an efficient inspection solution that could both rapidly screen large areas and accurately diagnose faulty areas.

[0035] In related technologies, fault detection of photovoltaic panels can be achieved by using a single drone equipped with high-precision inspection equipment to conduct full-site inspections along a pre-set route. The following describes a scenario using the related technology's drone inspection method for photovoltaic stations.

[0036] A photovoltaic power station employed a single drone inspection solution, employing a large drone equipped with a high-precision camera. While this drone offered high accuracy, its flight speed was limited, limiting its ability to inspect only 300 mu (approximately 16 acres) per day. Furthermore, due to the standardized inspection standards, even obvious surface defects required detailed scanning, significantly reducing inspection efficiency. In actual operation, battery depletion frequently interrupted missions, leading to frequent adjustments to inspection plans. Furthermore, due to its lack of intelligent planning capabilities, the drone often required multiple re-flights to complete inspections of suspicious areas, resulting in significant waste of time and energy.

[0037] The PV station drone inspection method described in the embodiments of this application uses a hierarchical inspection strategy that uses light drones for rapid initial screening and standard drones for precise re-inspection, achieving a balance between inspection efficiency and accuracy. This not only improves inspection efficiency but also ensures inspection quality in key areas. The following describes scenarios using the PV station drone inspection method described in this application.

[0038] A photovoltaic power station has implemented a dual-drone collaborative inspection system, using a light drone for rapid initial screening and a standard drone for precise re-inspection. The light drone scans the entire site at a speed of 8 meters per second, quickly marking suspicious areas. The system analyzes inspection data in real time and generates precise re-inspection routes. The standard drone then conducts detailed inspections of key areas using optimized timing and routes.

[0039] It can be seen that the drone inspection method for photovoltaic stations in the embodiment of the present application can not only achieve efficient inspection, but also effectively solve the time coordination and energy management problems in traditional single-machine inspection, thereby achieving the optimal configuration of inspection resources.

[0040] For ease of understanding, the following describes the process of the method provided by this implementation in combination with the above scenario. Figure 1 , is a flow chart of the drone inspection method for photovoltaic stations in an embodiment of the present application.

[0041] S101. Control the light UAV to perform inspections along a preset inspection route, and obtain real-time inspection images and actual flight parameters.

[0042] Light UAVs are small drones equipped with basic inspection equipment. These drones have fewer onboard devices than standard drones, are lightweight, and have longer flight times. Preset inspection routes are pre-planned inspection routes based on the layout of PV sites, including parameters such as the inspection area, flight altitude, and speed. Real-time inspection images are visible light images of PV panels captured by the drone during flight. Actual flight parameters represent the actual flight data, including speed, altitude, and heading.

[0043] The site inspection system performs this step at the start of a PV site inspection. Specifically, the system first generates a pre-defined inspection route covering the entire site based on the site's layout characteristics and inspection requirements. It then controls a light drone to take off and conduct a preliminary inspection along this route, capturing inspection images and flight parameter data in real time during flight. The inspection images are used for subsequent fault detection and analysis, while the flight parameters are used to assess actual inspection efficiency and optimize re-inspection planning.

[0044] In some embodiments, inspection control and data collection of light UAVs can be achieved in a variety of ways: Optionally, the station inspection system can adopt an automatic flight control method, automatically plan the flight path according to the preset inspection route, maintain stable flight through GPS positioning and attitude control system, and collect inspection images in real time through the onboard camera, and transmit the images and flight parameters to the ground station through the data link. Optionally, the station inspection system can also adopt a manual assisted control method, and the operator adjusts the flight attitude and detection parameters according to the actual situation to ensure that clear inspection images are obtained. It is understandable that other methods can also be used to achieve inspection control of light UAVs, such as using a hybrid control mode or multi-machine collaborative control.

[0045] In practice, it's crucial to ensure that light drones capture high-quality inspection images during high-speed inspections. To address this, the station inspection system employs an adaptive exposure control strategy: This strategy adjusts camera parameters in real time based on light intensity, maximizing flight speed while maintaining image clarity. It also incorporates an image stabilization algorithm to eliminate the effects of flight jitter. Furthermore, overlapping imaging ensures that no inspection areas are missed. These measures improve inspection efficiency while maintaining quality.

[0046] S102: Determine the real-time inspection progress and the fault detection area according to the real-time inspection image, and determine the predicted inspection time for the light UAV to complete the inspection based on the real-time inspection progress.

[0047] The real-time inspection progress indicates the percentage of light drones completing pre-set inspection routes. The fault detection area indicates the location of suspected faulty or abnormal PV panels detected through image analysis. The predicted inspection duration indicates the estimated time required to complete the remaining inspection tasks based on the current inspection speed.

[0048] The station inspection system performs this step in real time after acquiring real-time inspection images. Specifically, the system first matches the acquired inspection images with the preset inspection routes, calculates the percentage of areas where inspections have been completed, and obtains the real-time inspection progress. Simultaneously, the inspection images are analyzed in real time to identify potential fault features such as hot spots and cracks, marking the corresponding areas as fault detection areas. The system then calculates the predicted inspection duration based on the actual time spent in the completed areas and the area of ​​the remaining areas, combined with the current inspection speed.

[0049] In some embodiments, inspection progress and fault area determination can be achieved through a variety of methods: Optionally, the site inspection system can use image registration technology to match real-time inspection images with the PV site layout, calculate the image coverage area, and thus accurately determine the inspection progress. Simultaneously, a deep learning model can be used to analyze images in real time, automatically identifying various fault characteristics and marking the identification results on the site map. Optionally, the system can also combine historical inspection data to establish a fault prediction model, pre-estimating potential problem areas and improving inspection efficiency.

[0050] In practical applications, accurate predictions of the remaining inspection time are required. To address this, the station inspection system employs an adaptive prediction algorithm. First, a speed prediction model is established based on the actual inspection speed of completed areas. Then, environmental factors such as weather and time of day are considered to influence speed. Finally, a comprehensive calculation of the predicted duration is made based on the terrain characteristics and inspection difficulty of the remaining areas. This approach provides more accurate time estimates, facilitating the planning of subsequent re-inspection tasks.

[0051] S103: Determine multiple target areas to be inspected including historical fault areas and fault detection areas, and construct a real-time re-inspection route including the multiple target areas to be inspected.

[0052] The "historical fault area" indicates the area where faults were discovered during previous inspections. The "target inspection area" indicates the key areas that require re-inspection. The "real-time re-inspection route" indicates the optimized flight path connecting the target inspection areas.

[0053] The station inspection system performs this step after determining the fault detection area. Specifically, it first extracts historical fault area information from the historical inspection database and merges it with the fault detection area discovered during the current inspection to obtain a complete set of target inspection areas. Then, based on the spatial distribution characteristics of these areas, it uses a path planning algorithm to construct an optimal re-inspection route, ensuring that all target inspection areas are covered by the shortest path.

[0054] In some embodiments, the identification of target inspection areas and flight path planning can be achieved through a variety of methods: Optionally, the system can prioritize target inspection areas based on factors such as fault severity and duration, ensuring that critical fault areas are re-inspected first. Simultaneously, an improved ant colony algorithm can be used to calculate the optimal flight path, taking into account factors such as flight distance and number of turns. Optionally, the system can also use weather forecast data to plan re-inspection periods in advance to avoid adverse weather conditions.

[0055] In practice, it's necessary to balance the length of the re-inspection path with the completeness of the inspection. To address this, the station inspection system employs a block optimization strategy: First, adjacent target inspection areas are merged into inspection blocks to reduce duplicate flights; then, the optimal connection path is planned between the inspection blocks; and finally, the inspection sequence is dynamically adjusted based on actual conditions, ensuring inspection quality while improving efficiency.

[0056] S104. Calculate the equivalent flight parameters of a standard UAV during inspection under the same environmental conditions based on the actual flight parameters of the light UAV during inspection.

[0057] Actual flight parameters represent the speed, heading, altitude, and other operational data of a light UAV during an inspection. Equivalent flight parameters represent the expected flight performance of a standard UAV under the same conditions. Identical environmental conditions include external factors that affect flight, such as wind speed, temperature, and light.

[0058] The station inspection system performs this step after obtaining the actual flight parameters of the light UAV. Specifically, the system first establishes dynamic models for both aircraft types, including performance parameters such as thrust-to-weight ratio, payload capacity, and flight time. It then analyzes the actual flight data of the light UAV and, based on current environmental conditions, calculates the equivalent flight parameters of a standard UAV under the same conditions, including cruising speed, climb rate, and turning radius.

[0059] In some embodiments, flight parameter conversion can be achieved through a variety of methods: Optionally, the system can establish a mathematical model based on fluid dynamics principles that takes into account differences in aircraft weight, aerodynamic characteristics, and powertrain, and obtain accurate parameter mapping relationships through numerical calculations. Simultaneously, the model parameters can be corrected in real time based on historical flight data to improve conversion accuracy. Optionally, the system can also employ machine learning methods to establish performance correspondences between the two aircraft models through training with large amounts of flight data.

[0060] S105: Calculate the predicted re-inspection time required to complete a standard UAV inspection on a real-time re-inspection route based on the regional locations and equivalent flight parameters of the multiple target areas to be inspected.

[0061] The area location represents the spatial coordinates of the target inspection area within the PV plant. The predicted re-inspection duration represents the estimated time required for a standard drone to complete inspections of all target inspection areas.

[0062] The station inspection system performs this step after obtaining equivalent flight parameters. Specifically, the system first breaks down the real-time re-inspection route into an inspection segment and a transition segment. For the inspection segment, the required inspection time is calculated based on the area and inspection accuracy requirements. For the transition segment, the interval flight time is calculated based on the equivalent flight parameters. Finally, all these times are accumulated to obtain the predicted re-inspection duration.

[0063] In some embodiments, re-inspection duration prediction can be achieved through various methods: Optionally, the system can establish a refined time model that takes into account equipment startup and shutdown times, image acquisition time, and data processing time, and use simulation calculations to estimate the duration of each link. This also takes into account necessary interruptions such as battery replacement and equipment inspection. Optionally, the system can also establish a statistical model for time prediction based on historical re-inspection data to improve prediction accuracy.

[0064] During actual execution, unexpected delays must be addressed. To address this, the station inspection system employs a real-time correction mechanism: This system monitors the operating status of standard drones in real time and compares the deviation between actual and predicted times. When the deviation exceeds a threshold, it automatically adjusts the subsequent inspection plan to ensure overall time control. If necessary, it can dynamically adjust the inspection sequence or streamline the inspection process to ensure quality inspections in key areas.

[0065] S106: When the predicted re-inspection duration is greater than or equal to the predicted patrol inspection duration, the standard UAV is controlled to patrol according to the real-time re-inspection route to obtain a patrol inspection image.

[0066] The predicted re-inspection duration indicates the estimated time it will take a standard drone to complete the re-inspection task. The predicted patrol inspection duration indicates the estimated time it will take a light drone to complete the remaining patrol inspection tasks. The patrol inspection review images represent high-precision inspection images captured by a standard drone.

[0067] The station inspection system executes this step after confirming that the time conditions are met. Specifically, the station inspection system first compares the predicted re-inspection duration with the predicted inspection duration to ensure that the standard drone's re-inspection mission does not conflict with the light drone's initial inspection mission. When the time conditions are met, the standard drone is controlled to take off and conduct an inspection according to the real-time re-inspection route. During the re-inspection process, the onboard high-precision inspection equipment performs a detailed scan of the target inspection area to obtain a clear image of the fault characteristics.

[0068] In some embodiments, re-inspection control of standard drones can be achieved through various methods: Optionally, the system can employ adaptive flight control strategies, dynamically adjusting flight altitude and speed based on the characteristics of the inspection object to ensure image acquisition quality; while also analyzing image clarity in real time and automatically repeating acquisitions when necessary. Optionally, the system can also incorporate multi-sensor data fusion technology to simultaneously capture multiple images, including visible light and infrared, to provide more comprehensive fault information.

[0069] In practical applications, ensuring the safe operation of two drones in coordinated operations is crucial. To address this, the station inspection system employs multiple safety mechanisms: First, a real-time sharing mechanism for location information between the two drones is established; then, a minimum safe distance threshold is set, automatically triggering an avoidance procedure when the distance falls below it; Furthermore, a manual monitoring mechanism is implemented at the ground station, allowing for immediate control of the drones for emergency response. This approach effectively mitigates collision risks and ensures the safety of inspection operations.

[0070] S107: Generate an inspection report based on the real-time inspection image and the inspection review image.

[0071] Inspection reports are comprehensive analysis documents containing information such as fault type, location, and severity. Real-time inspection images represent preliminary screening images captured by light drones. Inspection review images represent detailed inspection images captured by standard drones.

[0072] The station inspection system performs this step after completing all inspection tasks. Specifically, the system first matches and compares the two images to confirm the consistency of the fault characteristics. It then uses image recognition algorithms to extract the fault characteristics and determine the fault type and severity. Finally, the analysis results are compiled into an inspection report in a pre-set format, including a description of the fault, location information, and recommended solutions.

[0073] In some embodiments, inspection reports can be generated in a variety of ways: Optionally, the system can employ intelligent diagnostic algorithms, using deep learning models to automatically identify and classify various fault types, generating standardized fault descriptions. This system can also incorporate historical maintenance data to provide targeted treatment recommendations for each fault. Optionally, the system can also establish a fault impact assessment model to calculate the impact of each fault on power generation efficiency and assist in prioritizing repairs.

[0074] In practice, ensuring the accuracy and reliability of reports is crucial. To address this, the station inspection system employs a multi-level verification mechanism: First, an algorithm assesses the credibility of fault identification results; then, professionals review and confirm critical faults; and finally, a report review process is established to ensure the accuracy of output results. The system also saves all original image data for subsequent verification and analysis. This approach significantly improves the reliability of inspection reports and provides a solid basis for subsequent maintenance work.

[0075] The following is a more detailed description of the process of the method provided by this implementation. Figure 2 , is another flow chart of the drone inspection method for photovoltaic stations in an embodiment of the present application.

[0076] S201. Control the light UAV to conduct inspections along a preset inspection route, and obtain real-time inspection images and actual flight parameters.

[0077] Referring to step S101, the station inspection system will control the light UAV to perform an initial inspection.

[0078] S202: Determine the real-time inspection progress and the fault detection area according to the real-time inspection image, and determine the predicted inspection time for the light UAV to complete the inspection based on the real-time inspection progress.

[0079] Referring to step S102, the station inspection system will determine the fault detection area and predicted inspection duration of the light UAV.

[0080] S203: Determine multiple target areas to be inspected including historical fault areas and fault-detected areas, and construct a real-time re-inspection route including the multiple target areas to be inspected.

[0081] Referring to step S103 , the station inspection system will construct a real-time re-inspection route.

[0082] S204. Based on the power parameter configurations of the light UAV and the standard UAV, construct a flight operation mapping of the two models when responding to the same wind conditions.

[0083] The power parameter configuration represents the drone's performance parameters, such as thrust-to-weight ratio, maximum takeoff weight, and battery capacity. The flight operation mapping represents the corresponding flight performance of two aircraft models under the same wind conditions. The wind environment represents meteorological conditions that affect flight, including wind speed and direction.

[0084] The station inspection system performs this step before calculating equivalent flight parameters. Specifically, the system first collects detailed power parameters for both aircraft models, including motor power, propeller characteristics, and airframe weight. It then develops a mathematical model that accounts for aerodynamic characteristics and analyzes the forces under different wind conditions. Finally, through extensive experimental data verification, a performance mapping relationship between the two aircraft models under various wind conditions is obtained.

[0085] In some embodiments, the flight operations mapping can be constructed using a variety of methods: Optionally, the system can employ computational fluid dynamics to simulate the aerodynamic characteristics of two aircraft models under different wind speeds and wind directions, creating detailed performance curves. This model can also be calibrated with actual flight data to improve mapping accuracy. Alternatively, the system can utilize machine learning algorithms to analyze historical flight data to develop an adaptive mapping model, namely a flight performance mapping model.

[0086] The mathematical processing logic of the flight performance mapping model revolves around the conversion of flight characteristics between two different aircraft models under similar environmental conditions. Based on the principles of fluid mechanics, the model constructs a set of mathematical equations that take into account differences in thrust-to-weight ratio, aerodynamic characteristics, and powertrain systems. In dynamic modeling, the drone is simplified into a point mass model, and a set of differential equations describing the flight state is established using Newton's laws of motion. The model inputs include environmental parameters such as wind speed and temperature, as well as real-time flight data (speed, altitude, attitude angle, etc.) of the light drone. By solving the state equations, the equivalent flight parameters of a standard drone are derived. For example, when a light drone is flying steadily at 8 m / s in a 5 m / s crosswind, the model calculates the thrust compensation and attitude adjustment parameters required for the standard drone based on the differences in mass ratio and aerodynamic characteristics between the two models to achieve the same flight stability.

[0087] In engineering practice, it's necessary to handle extreme weather conditions such as sudden strong winds. To address this, the station inspection system employs a segmented mapping strategy: First, wind conditions are classified by intensity, with independent mapping relationships established for each level. A safety margin is then set to automatically reduce flight parameters when wind speeds exceed a threshold. An emergency avoidance mechanism is also established to ensure timely adjustment of flight status in extreme situations. This approach improves the system's adaptability to various weather conditions.

[0088] S205. Based on the inspection parameter configurations of the light UAV and the standard UAV, a detection time mapping is constructed for the two types of UAVs when processing the same inspection object.

[0089] Inspection parameter configuration represents device parameters such as camera resolution, sampling frequency, and detection accuracy. Inspection time mapping represents the time required for two models to complete the same inspection task. Inspection objects represent the PV panels or equipment to be inspected.

[0090] The inspection system performs this step after constructing the flight operation mapping. Specifically, the system first compares the performance differences between the inspection equipment of the two aircraft models, including image acquisition speed and data processing capabilities. It then analyzes the technical requirements of different inspection tasks and establishes a correlation model between inspection accuracy and time. Finally, through actual testing and verification, it obtains a time mapping relationship between the two aircraft models in various inspection scenarios.

[0091] In some embodiments, the construction of the detection time map can be achieved through various methods: Optionally, the system can establish a time prediction model based on image quality, dynamically adjust detection parameters based on target feature complexity, and optimize acquisition efficiency; while also considering data transmission and processing delays to achieve more accurate time estimation. Optionally, the system can also use an intelligent scheduling algorithm to dynamically allocate time resources based on the priority of the detection task.

[0092] In practical applications, it's necessary to balance inspection efficiency and quality requirements. To address this, the station inspection system employs a multi-level inspection strategy: Different detection accuracy requirements are set based on fault type, with detailed scanning of key areas and rapid inspection of general areas. A quality assessment mechanism is also established, automatically increasing inspection time when image quality falls short. This approach ensures inspection quality while improving overall efficiency.

[0093] S206: Convert the actual flight parameters of the light UAV into equivalent flight parameters of the standard UAV according to the flight operation mapping and the detection time mapping.

[0094] Actual flight parameters represent data such as speed, altitude, and power consumption recorded during inspections for light drones. Equivalent flight parameters represent the equivalent flight parameters for standard drones after conversion. Mapping conversion involves converting the parameters of one aircraft model into those of another through a corresponding relationship.

[0095] The inspection system performs this step after obtaining the two mapping relationships. Specifically, the inspection system first converts the actual flight parameters of the light UAV according to the flight operation mapping to obtain the baseline flight parameters of the standard UAV. It then adjusts these baseline parameters based on the test time mapping, accounting for time losses caused by equipment performance differences. Finally, it comprehensively considers the impact of wind conditions and test requirements to obtain the final equivalent flight parameters.

[0096] In some embodiments, parameter conversion can be achieved through a variety of methods: Optionally, the system can establish a multi-factor coupled conversion model, taking into account the influence of environmental factors such as wind speed, temperature, and sunlight to achieve more accurate parameter mapping; At the same time, an adaptive correction mechanism can be introduced to continuously optimize conversion accuracy based on real-time feedback. Optionally, the system can also adopt a data-driven approach, using machine learning algorithms to mine more complex parameter relationships from historical data.

[0097] In actual implementation, it's necessary to address the cumulative error in parameter conversion. To address this, the station inspection system employs a segmented calibration strategy: First, the flight process is divided into multiple characteristic segments, and parameters are reset at the start of each segment. Then, by monitoring the error trend in real time, a calibration procedure is triggered when the cumulative error exceeds a threshold. A parameter anomaly detection mechanism is also established to promptly identify and correct unreasonable conversion results. This approach effectively controls error propagation and improves the reliability of conversion results.

[0098] S207 : Based on the regional locations of the multiple target areas to be inspected, the real-time re-inspection route is divided into an inspection section and a transition section.

[0099] The target inspection area represents the PV panel area that requires focused inspection. The inspection section represents the route segment within the target inspection area where detailed inspections are performed. The transition section represents the transfer route segment connecting adjacent target inspection areas.

[0100] The terminal inspection system performs this step after determining the equivalent flight parameters. Specifically, it first analyzes the spatial distribution of the target inspection area to determine the positional relationships and inspection boundaries of each area. It then plans inspection paths within each area based on the inspection requirements, forming inspection sections. Finally, based on the relative positions of the areas, it designs optimal connecting paths to form transition sections.

[0101] In some embodiments, route segmentation can be achieved through a variety of methods: Optionally, the system can employ an intelligent zoning algorithm to cluster inspection areas based on fault type and severity, optimizing the division of inspection segments. Simultaneously, a path planning algorithm can be used to generate the shortest transition paths, reducing inefficient flight time. Optionally, the system can also consider terrain obstacles and airflow conditions to design safer flight paths for transition segments.

[0102] In engineering applications, it's necessary to address the overlap of adjacent inspection areas. To address this, the station inspection system employs a fusion optimization strategy: first, the boundaries of the overlapping areas are identified and divided into inspection sections that are most suitable for inspection. The inspection sequence of adjacent sections is then optimized to avoid duplicate inspections. Furthermore, transitional sections are designed with room for maneuver to adjust the inspection range based on actual conditions. This approach improves inspection integrity and efficiency.

[0103] S208. Calculate the inspection time of the standard UAV in the inspection section based on the equivalent flight parameters.

[0104] The inspection segment duration represents the time it takes a standard drone to complete a detailed inspection of a target area. The equivalent flight parameters represent the actual operating parameters of a standard drone after accounting for environmental factors. The inspection path represents the specific flight trajectory within the target area.

[0105] The terminal inspection system performs this step after completing route segmentation. Specifically, the system first calculates the inspection path length required for complete coverage based on the area and shape of the target inspection area. It then calculates the pure inspection time, combining the speed and inspection interval requirements from the equivalent flight parameters. Finally, it considers the time required for auxiliary actions such as steering and attitude adjustments to determine the total time for each inspection segment.

[0106] In some embodiments, segment time calculation can be implemented in a variety of ways: Optionally, the system can establish a hierarchical time model that takes into account detection accuracy requirements, employing different inspection densities for different fault areas to achieve differentiated inspection time allocation; Furthermore, image quality assessment results can be combined to dynamically adjust inspection parameters and flight speed. Optionally, the system can also incorporate a parallel processing mechanism to simultaneously complete data processing and analysis during flight.

[0107] In practice, unexpected situations need to be handled during inspections. To address this, the station inspection system employs a flexible time management strategy: First, an emergency time margin is reserved in time calculations; then, a real-time monitoring mechanism is established to automatically trigger repeated inspections when inspection quality falls short of standards; and a maximum inspection time limit is set to ensure that no single section takes up too much time. This approach improves the reliability and timeliness of inspections.

[0108] S209. Calculate the transition section time of the standard UAV in the transition section based on the equivalent flight parameters and the power constraints of the standard UAV.

[0109] The transition duration represents the time required for a standard drone to move between adjacent target inspection areas. Power constraints include restrictions such as battery life and maximum speed. The navigation path represents the specific flight trajectory within the transition segment.

[0110] The station inspection system performs this step after calculating the inspection segment duration. Specifically, the system first determines the optimal transition path based on the positional relationship between adjacent target inspection areas. It then calculates the maximum achievable flight speed, taking into account the power constraints of a standard drone. Finally, it calculates the expected duration for each transition segment, combining the path length and speed parameters.

[0111] In some embodiments, transition time calculation can be achieved through a variety of methods: Optionally, the system can use a dynamic path planning algorithm to select the optimal flight path based on real-time wind field data, taking into account the time consumed during acceleration and deceleration. Furthermore, the system can monitor the battery charge status and plan a transfer route to a charging station when necessary. Optionally, the system can also establish an energy-based time estimation model to optimize the speed configuration.

[0112] In engineering practice, it's necessary to balance flight speed and energy consumption. To address this, the station inspection system employs an energy efficiency optimization strategy: First, a speed-energy consumption relationship model is established to identify the optimal operating point; then, flight speed in transitional sections is dynamically adjusted based on remaining battery power; wind direction is also considered, utilizing tailwind conditions to improve flight efficiency. This approach ensures timely delivery while extending flight time.

[0113] In some embodiments, after inspecting a target area to be inspected, the station inspection system will perform power planning. If the power is insufficient, it will go to the battery exchange position for battery exchange. At this time, the transition section is a complete battery exchange path from the previous inspection section to the battery exchange position, and then from the battery exchange position to the next inspection section. The station inspection system will ensure the maximum flight speed in the transition section to perform the inspection of the next inspection section as soon as possible. In this case, the station inspection system will calculate the remaining power of the standard drone after the previous inspection section based on the battery capacity and power consumption characteristics of the standard drone; when the remaining power is lower than the preset power threshold, the battery exchange position is determined so that the standard drone performs the battery exchange operation before the next inspection section; based on the equivalent flight parameters and the power constraints of the standard drone, the maximum flight speed of the standard drone to and from the battery exchange position is determined, and the battery exchange path time is calculated; based on the battery exchange path time, the equipment start and stop time and the expected battery exchange time, the transition section time of the standard drone is calculated.

[0114] Among them, the battery capacity represents the rated energy storage of a standard drone power battery. The power consumption characteristics represent the energy consumption pattern under different operating conditions. The remaining power represents the remaining energy value in the current battery. The preset power threshold represents the minimum power standard that triggers the battery replacement operation. The battery replacement location represents the geographical coordinate point where the battery replacement is performed. The battery replacement route time represents the flight time required to return to the battery replacement location. The equipment start-stop time represents the operating time of the drone landing and taking off. The estimated battery replacement time represents the standard operating time for battery replacement.

[0115] The station inspection system performs this step after a standard drone completes an inspection section. Specifically, the station inspection system first establishes an accurate energy consumption prediction model based on the real-time current and voltage data recorded by the battery management system, combined with flight status parameters. This model then calculates the remaining battery charge and compares it with a preset safety threshold. When the remaining charge approaches the threshold, the station inspection system selects the optimal battery swap location from multiple preset battery swap stations, taking into account flight distance, weather conditions, and terrain characteristics. Finally, the round-trip time, equipment operating time, and standard battery swap time are calculated to determine the duration of the complete battery swap process.

[0116] It's important to note that the mathematical processing logic of the energy consumption prediction model focuses on analyzing battery discharge characteristics and estimating remaining charge. The model uses a state estimation algorithm based on a Kalman filter, combining observables such as battery voltage and current with a battery equivalent circuit model to estimate the battery's state of charge (SOC) in real time. In the prediction phase, the model uses time series analysis, combining the planned route distance and expected power consumption, to predict the remaining charge at a future point in time. The model also considers the impact of temperature on battery performance, correcting battery parameters using the Arrhenius equation. For example, when a standard drone performs a re-inspection mission, the model predicts the remaining charge at mission completion based on the current battery state (e.g., voltage of 22V, remaining charge of 80%) and mission parameters (e.g., expected flight distance of 2km, average power consumption of 200W). This prediction then determines whether a battery replacement should be scheduled in advance.

[0117] In some embodiments, the battery swap management strategy can be implemented in a variety of ways: Optionally, the station inspection system can establish an intelligent prediction mechanism to plan the timing of battery swaps in advance by analyzing historical energy consumption data, current weather conditions, and flight mission characteristics, including establishing a power consumption prediction model, formulating a multi-objective battery swap station selection strategy, and designing an optimal path planning solution. Optionally, the station inspection system can also adopt a real-time optimization strategy to dynamically adjust the battery swap plan by continuously monitoring the battery status, environmental changes, and mission progress, including real-time evaluation of the remaining range, calculation of energy margin, and optimization of the battery swap station layout. It is understandable that other methods can also be used to implement battery swap management, such as a distributed battery swap station network or a mobile battery swap platform. In addition, it is necessary to supplement the emergency response mechanism during the battery swap process to ensure battery safety in extreme situations.

[0118] In practical applications, it's necessary to handle unexpected weather changes during battery swapping. To address this, the station inspection system employs an adaptive battery swapping strategy: First, a multi-layered weather monitoring network is established to collect real-time meteorological data such as wind speed and precipitation. Then, battery swapping windows are adjusted in advance based on weather forecasts. In the event of inclement weather, the system automatically selects the nearest indoor battery swapping station and calculates a safe backup route. Furthermore, battery temperature is monitored in real time, initiating a forced cooling process when necessary. This solution significantly improves the safety and reliability of battery swapping operations.

[0119] S210: Accumulate the inspection section time and the transition section time to obtain the predicted re-inspection time after the standard UAV inspection is completed.

[0120] The predicted re-inspection duration represents the total expected time for a standard drone to complete inspections of all target areas. Accumulation represents the process of adding up the times for each section in the order they are executed. The time margin represents the buffer time reserved to account for uncertainties.

[0121] The station inspection system performs this step after obtaining the time for all sections. Specifically, the system first adds up the time for each inspection section and transition section in the inspection order. It then adds an appropriate time margin to account for uncertainties such as weather changes and equipment commissioning. Finally, it obtains the total predicted re-inspection duration, which includes both primary and secondary operation times.

[0122] In some embodiments, total duration prediction can be achieved through a variety of methods: Optionally, the system can establish a probabilistic model that considers time correlation, analyze the fluctuation patterns of time consumption in each segment, and obtain a more accurate total duration distribution; at the same time, a reasonable time margin ratio can be determined based on historical data. Optionally, the system can also use Monte Carlo simulation methods to obtain more reliable time prediction results through multiple simulations.

[0123] In practical applications, time prediction accuracy needs to be improved. To address this, the station inspection system employs a progressive correction mechanism: It first generates a rough prediction based on initial data; then, as the inspection progresses, it continuously collects actual time data; and finally, through real-time comparison, it dynamically adjusts the time allocation for subsequent sections. This approach improves the accuracy of predictions and provides a more reliable basis for task scheduling.

[0124] S211. When the predicted re-inspection duration is greater than or equal to the predicted patrol inspection duration, the standard UAV is controlled to patrol according to the real-time re-inspection route to obtain a patrol inspection image.

[0125] Referring to step S106, the station inspection system will control the standard drone to re-inspect.

[0126] In some embodiments, since the target area to be inspected includes historical fault areas and fault detection areas, and the standard drone only performs inspections of the target area to be inspected, the maximum speed flight transition will be guaranteed for other locations on the route; at this time, after inspecting the fault detection area, since the light drone has not yet completed its inspection, the remaining historical fault areas of the standard drone may be located on the front route of the light drone, which will cause a conflict in the routes of the two drones. In this case, the station inspection system will calculate the re-inspection time nodes when the standard drone arrives at multiple target areas to be inspected when inspecting according to the real-time re-inspection route; calculate the inspection time nodes when the light drone arrives at multiple target areas to be inspected based on the real-time inspection progress; when the re-inspection time node is earlier than the inspection time node, the corresponding target area to be inspected will be marked as a unit overlap area; determine the re-inspection takeoff delay time so that the re-inspection time node of the unit overlap area is later than the inspection time node.

[0127] The re-inspection time node represents the estimated time when a standard drone will arrive at each target inspection area as planned. The inspection time node represents the estimated time when a light drone will complete its initial inspection of the corresponding area. The crew overlap area represents an area where two drones may operate simultaneously during the same time period. The re-inspection takeoff delay time represents the delay required to avoid interference. The time node calculation represents the process of estimating the arrival time based on flight speed and path length.

[0128] The station inspection system performs this step before planning the standard drone re-inspection mission. Specifically, the station inspection system first calculates the arrival time of the standard drone at each target inspection area based on the real-time re-inspection route and equivalent flight parameters. It then predicts the completion time of each area based on the light drone's real-time inspection progress and remaining path. If the re-inspection time for certain areas is found to be earlier than the initial inspection time, the station inspection system marks these areas as overlapping and adjusts the standard drone's takeoff time to ensure the rationality of the two aircraft's operation sequence.

[0129] In some embodiments, coordinated management of time nodes can be achieved through a variety of methods: Optionally, the station inspection system can employ dynamic programming methods to precisely schedule the operations of the two aircraft by establishing a spatiotemporal conflict detection model, calculating an optimal delay strategy, and generating a coordinated time table. Optionally, the station inspection system can also implement adaptive time management to ensure operational safety through real-time monitoring of the positions of the two aircraft, predicting potential conflicts, and dynamically adjusting flight speeds. It is understood that other methods can also be used to achieve time node coordination, such as alternating operations in different areas or setting up safety buffer zones. Furthermore, additional explanation is needed regarding the priority handling mechanism in emergency situations.

[0130] In engineering practice, it's necessary to address time node deviations caused by weather changes. To address this, the station inspection system employs a multi-level time compensation mechanism: First, a flight time correction model based on meteorological data is established to assess the impact of weather on flight speed in real time. Flexible buffer times are then set between inspection areas to absorb delays caused by uncertainties. When significant time deviations occur, the system reschedules subsequent inspections to ensure overall task coordination. This solution effectively addresses scheduling needs under complex weather conditions.

[0131] In some embodiments, incremental supplementation of the re-inspection area can be performed, that is, the station inspection system will determine one or more new areas to be inspected based on the real-time inspection image on the preceding route corresponding to the unit overlapping area on the real-time re-inspection route; and recalculate the re-inspection time node of the unit overlapping area based on the newly added areas to be inspected, so that the re-inspection time node of the unit overlapping area is later than the inspection time node.

[0132] The preceding route represents the flight path preceding the crew overlap area. Newly inspected areas represent areas of particular concern discovered during the initial inspection. Time node recalculation represents the process of adjusting the re-inspection plan to account for the newly added areas. Inspection sequence optimization involves rearranging the inspection route based on time constraints.

[0133] The station inspection system performs this step after discovering overlapping areas. Specifically, it first analyzes real-time inspection images captured by the light drone on its preceding route to identify areas with anomalies or requiring further verification. These newly inspected areas are then added to the standard drone's re-inspection task queue, and the inspection time for each area is recalculated. By rationally arranging the inspection order for the newly added areas, the inspection time for overlapping areas is naturally delayed.

[0134] In some embodiments, dynamic adjustment of the detection area can be achieved in a variety of ways: Optionally, the station inspection system can adopt a real-time image analysis strategy to quickly identify the newly added inspection areas by performing steps such as image preprocessing, feature extraction, anomaly detection, area positioning, and priority assessment; Optionally, the station inspection system can also implement an intelligent planning method to ensure the reasonable arrangement of the newly added areas by analyzing task dependencies, calculating time windows, optimizing insertion positions, adjusting flight paths, and verifying feasibility. It is understandable that other methods can also be used to achieve dynamic adjustment of the detection area, such as task reorganization based on multi-objective optimization or adaptive task decomposition. In addition, the quality assurance mechanism for the detection of newly added areas needs to be supplemented.

[0135] In practical applications, it's necessary to ensure inspection integrity while avoiding excessive delays. To address this, the station inspection system employs a hierarchical processing strategy: First, the urgency of newly added areas to be inspected is assessed, establishing a priority model based on failure risk. The optimal inspection time window for each newly added area is then calculated, and the new tasks are inserted into the appropriate time period without affecting inspections in existing key areas. When the cumulative delay exceeds a threshold, the system initiates a task optimization program, compressing inspection time by adjusting inspection accuracy or merging adjacent areas. This approach maintains the effectiveness of inspection plans in dynamic environments.

[0136] S212: Generate an inspection report based on the real-time inspection image and the inspection review image.

[0137] Referring to step S107 , the station inspection system generates an inspection report.

[0138] In some embodiments, the station inspection system will generate a corresponding maintenance task reference list after generating an inspection report to ensure the normal operation of the photovoltaic station, that is, the station inspection system will extract fault image features from the inspection review image to form a fault list; match the fault types of multiple detected faults in the fault list with the power generation impact parameter table preset by the photovoltaic station to obtain the fault impact degree value of the detected fault; sort the detected faults according to the fault impact degree value to obtain a task list of maintenance tasks corresponding to the detected faults; generate fault location coordinates and maintenance instructions corresponding to multiple maintenance tasks in the task list.

[0139] Among them, the fault image feature represents the visual feature information reflecting the abnormal status of the equipment. The fault list represents a structured dataset containing information such as fault type, location, and characteristic parameters. The power generation impact parameter table provides a quantitative indicator of the impact of different types of faults on photovoltaic power generation efficiency. The fault impact value represents the impact assessment result of a specific fault on system operation. The maintenance task represents the treatment plan developed for a specific fault. The maintenance instructions provide specific operational instructions including maintenance methods, required tools, precautions, etc.

[0140] The station inspection system performs this step after obtaining complete inspection data. Specifically, the system first conducts in-depth analysis of the inspection review images to extract characteristic information for various fault types, including appearance characteristics, temperature distribution, and material changes. These characteristics are then matched against a pre-established fault type library to determine the specific fault type. Based on a table of power generation impact parameters, the system assesses the impact of each fault and ranks them by severity. Finally, based on the fault characteristics and maintenance experience, a detailed repair guide is generated.

[0141] In some embodiments, fault analysis and report generation can be achieved in a variety of ways: Optionally, the site inspection system can use intelligent diagnostic methods to accurately identify and evaluate faults by performing image enhancement processing, multimodal feature fusion, deep learning classification, fault location mapping, impact assessment modeling, and other steps; Optionally, the site inspection system can also implement a knowledge-driven analysis strategy to form a complete maintenance solution by building a fault knowledge graph, extracting expert experience rules, generating a processing decision tree, formulating maintenance strategies, and compiling work guides. It is understandable that other methods can also be used to achieve fault analysis and report generation, such as fault prediction based on statistical models or adaptive maintenance plan generation. In addition, the quality verification and update mechanism of the report needs to be supplemented.

[0142] In practice, it's necessary to handle complex situations where multiple faults coexist and influence each other. To address this, the station inspection system employs a systematic analysis approach: first, a fault correlation model is established to analyze the causal relationships and impact paths between different faults. Then, graph theory algorithms are used to determine the fault propagation chain and identify the root cause. When generating maintenance tasks, the system considers the repair dependencies between faults and arranges the repair sequence appropriately. Furthermore, the system predicts the improvement in repair effectiveness after repair, providing a decision-making basis for allocating maintenance resources. This approach improves the efficiency and accuracy of complex fault handling.

[0143] In the embodiment of the present application, due to the use of a hierarchical inspection mode in which light UAVs and standard UAVs work together, and by establishing a flight operation mapping and detection time mapping relationship between the two types of aircraft, accurate parameter conversion and timing coordination based on real-time data are achieved, so that the inspection efficiency can be significantly improved under the premise of ensuring the inspection quality. Through technical means such as dynamic power management, multi-region collaborative scheduling, and real-time task optimization, the technical problems of the traditional single-machine inspection, such as the difficulty in balancing detection accuracy and efficiency, the untimely response to sudden failures, and the poor adaptability to severe weather, are effectively solved, thereby achieving the high efficiency, high quality, and high reliability goals of photovoltaic power station inspections. Especially in complex environments, this application successfully overcomes various challenges in the multi-machine collaboration process through innovative technologies such as intelligent battery replacement strategies, time node coordination mechanisms, dynamic adjustment methods for detection areas, and systematic fault analysis, thereby improving the overall performance and practical value of the inspection system and providing reliable technical support for the intelligent operation and maintenance of photovoltaic power stations.

[0144] The following describes the station inspection system in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , which is a schematic diagram of the physical device structure of the station inspection system in an embodiment of the present application.

[0145] It should be noted that Figure 3 The structure of the station inspection system shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0146] like Figure 3 As shown, the station inspection system includes a CPU 301, which can perform various appropriate actions and processes based on programs stored in a ROM 302 or programs loaded from a storage unit 308 into a RAM 303, such as executing the methods described in the above embodiments. RAM 303 also stores various programs and data required for system operation. CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An I / O interface 305 is also connected to bus 304.

[0147] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, push button switches, and the like; an output section 307 including a liquid crystal display (LCD), an audio output device, indicator lights, and the like; a storage section 308 including a hard disk and the like; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. Removable media 311, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 310 as needed, so that computer programs read from the removable media can be installed in the storage section 308 as needed.

[0148] In particular, according to an embodiment of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, an embodiment of the present invention includes a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 309 and / or installed from the removable medium 311. When the computer program is executed by the CPU 301, the various functions defined in the present invention are performed.

[0149] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings.

[0150] Specifically, the station inspection system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the drone inspection method for photovoltaic stations provided by the above embodiment is implemented.

[0151] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the station inspection system described in the above embodiments, or may exist independently and not be incorporated into the station inspection system. The storage medium carries one or more computer programs, which, when executed by a processor of the station inspection system, enable the station inspection system to implement the drone inspection method for photovoltaic stations provided in the above embodiments.

[0152] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0153] As used in the above embodiments, the term “when” may be interpreted to mean “if” or “after” or “in response to determining that” or “in response to detecting that”, depending on the context. Similarly, the phrases “upon determining that” or “if (stated condition or event) is detected” may be interpreted to mean “if determining that” or “in response to determining that” or “upon detecting (stated condition or event)” or “in response to detecting (stated condition or event)”, depending on the context.

Claims

1. A method for drone inspection of photovoltaic stations, characterized in that: Applied to a station inspection system, the station inspection system includes a light UAV and a standard UAV, the light UAV has fewer onboard devices than the standard UAV, and the light UAV weighs less than the standard UAV; the method includes: Controlling the light UAV to conduct inspections along a preset inspection route, obtaining real-time inspection images and actual flight parameters; the actual flight parameters include actual flight speed and heading adjustment data for each area segment; Determining a real-time inspection progress and a fault detection area based on the real-time inspection image, and determining a predicted inspection time for the light UAV to complete the inspection based on the real-time inspection progress; Determine multiple target inspection areas including the historical fault area and the fault detection area, and construct a real-time re-inspection route including the multiple target inspection areas; Calculate the equivalent flight parameters of the standard UAV during inspection under the same environmental conditions based on the actual flight parameters of the light UAV during inspection; Calculating a predicted re-inspection time for completing the standard UAV inspection on the real-time re-inspection route based on the regional locations of the multiple target areas to be inspected and the equivalent flight parameters; When the predicted re-inspection duration is greater than or equal to the predicted inspection duration, controlling the standard UAV to perform inspection according to the real-time re-inspection route to obtain an inspection review image; An inspection report is generated based on the real-time inspection image and the inspection review image.

2. The method according to claim 1, characterized in that The step of calculating the equivalent flight parameters of the standard UAV under the same environmental conditions based on the actual flight parameters of the light UAV during inspection specifically includes: Based on the power parameter configurations of the light UAV and the standard UAV, a flight operation mapping of the two models in the same wind environment is constructed; Based on the inspection parameter configurations of the light UAV and the standard UAV, a detection time mapping is constructed for the two types of UAVs when processing the same inspection object; The actual flight parameters of the light UAV are converted into equivalent flight parameters of the standard UAV according to the flight operation mapping and the detection time mapping.

3. The method according to claim 1, characterized in that The step of calculating the predicted re-inspection time required for completing the standard UAV inspection on the real-time re-inspection route based on the regional locations of the multiple target areas to be inspected and the equivalent flight parameters specifically includes: Based on the location of the multiple target areas to be inspected, the real-time re-inspection route is divided into a patrol section and a transition section; the patrol section is located in the target areas to be inspected, and the transition section is located between adjacent target areas to be inspected; Calculating the inspection time of the standard UAV in the inspection section according to the equivalent flight parameters; Calculating the time taken by the standard UAV in the transition section according to the equivalent flight parameters and the power constraints of the standard UAV; The inspection section time and the transition section time are accumulated to obtain the predicted re-inspection time after the standard UAV inspection is completed.

4. The method according to claim 3, characterized in that The step of calculating the time taken by the standard UAV in the transition section according to the equivalent flight parameters and the power constraints of the standard UAV specifically includes: Calculating the remaining power of the standard drone after the last inspection segment based on the battery capacity and power consumption characteristics of the standard drone; When the remaining power is lower than a preset power threshold, determining a battery replacement position so that the standard drone performs a battery replacement operation before the next inspection section; Determine the maximum flight speed of the standard UAV to and from the battery swap location based on the equivalent flight parameters and the power constraints of the standard UAV, and calculate the battery swap path time; The transition section time of the standard UAV is calculated based on the battery replacement path time, equipment start and stop time, and expected battery replacement time.

5. The method according to claim 1, wherein Before the step of controlling the standard drone to perform inspection according to the real-time review route and obtaining the inspection review image, the method further includes: Calculating the re-inspection time nodes when the standard UAV arrives at the multiple target inspection areas when performing inspection according to the real-time re-inspection route; Calculating the inspection time nodes when the light UAV arrives at the multiple target inspection areas according to the real-time inspection progress; When the re-inspection time node is earlier than the inspection time node, the corresponding target area to be inspected is marked as a unit overlap area; The re-inspection takeoff delay time is determined so that the re-inspection time node of the crew overlap area is later than the inspection time node.

6. The method according to claim 5, characterized in that When the re-inspection time node is earlier than the inspection time node, after the step of marking the corresponding target area to be inspected as a unit overlap area, the method further includes: On a preceding route corresponding to the overlapping area of ​​the unit on the real-time re-inspection route, determining one or more newly added areas to be inspected based on the real-time inspection image; The re-inspection time node of the unit overlapping area is recalculated based on the newly added area to be inspected, so that the re-inspection time node of the unit overlapping area is later than the inspection time node.

7. The method according to claim 1, characterized in that After the step of generating an inspection report based on the real-time inspection image and the inspection review image, the method further includes: Extracting fault image features from the inspection and review images to form a fault list; Matching the fault types of the multiple detected faults in the fault list with a power generation impact parameter table preset by the photovoltaic station to obtain a fault impact degree value of the detected fault; Sort the detected faults according to the fault impact values ​​to obtain a task list of maintenance tasks corresponding to the detected faults; Generate fault location coordinates and maintenance instructions corresponding to multiple maintenance tasks in the task list.

8. A station inspection system, characterized in that: The site inspection system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the site inspection system to execute the method described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instruction is executed on the station inspection system, the station inspection system is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a station inspection system, the station inspection system is enabled to execute the method according to any one of claims 1 to 7.

Images