Photovoltaic inspection management method and system based on unmanned aerial vehicle

Abstract

The invention relates to the technical field of photovoltaic power station inspection. The invention provides a photovoltaic inspection management method and system based on an unmanned aerial vehicle, and the method comprises the steps: constructing an inspection management database based on historical data; respectively constructing an inspection planning model and a data analysis and evaluation model based on the inspection management database; planning a flight path and an endurance strategy through the inspection planning model, and dynamically optimizing the flight path and the endurance strategy through the inspection planning model according to the real-time data; an inspection task is executed according to the dynamically optimized flight path and the endurance strategy, related data of the photovoltaic power station in the inspection process are collected in real time, and the related data are returned in real time; related data returned in real time is analyzed and evaluated through the data analysis and evaluation model, and corresponding operation and maintenance measures are formulated according to evaluation results, so that the potential problems that an existing inspection mode depends on manual inspection, the efficiency is low, each corner of the photovoltaic power station is difficult to comprehensively cover, and missing is likely to occur are solved.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic power station inspection technology, and in particular to a photovoltaic inspection management method and system based on unmanned aerial vehicle (UAV). Background Art

[0002] As the global demand for renewable energy continues to grow, photovoltaic power generation, as a clean and renewable energy form, has become increasingly important. However, the operation and maintenance management of photovoltaic power stations faces many challenges, such as large power station area, difficult inspection, high operation and maintenance costs, and delayed fault detection.

[0003] The existing inspection method relies on manual inspection, which is not only inefficient but also difficult to fully cover every corner of the photovoltaic power station and is prone to missing potential problems. Summary of the invention

[0004] The purpose of the present invention is to provide a photovoltaic power generation energy storage grid-connected optimization system and method and system, aiming to solve the problem that the existing inspection method relies on manual inspection, which is not only inefficient, but also difficult to fully cover every corner of the photovoltaic power station and easily misses potential problems.

[0005] The embodiments of the present invention are implemented by the following technical solutions:

[0006] A photovoltaic inspection management method based on drones, comprising the following steps:

[0007] Build an inspection management database based on historical data related to photovoltaic power plants and drone systems;

[0008] Based on the inspection management database, the inspection planning model and data analysis and evaluation model are constructed respectively;

[0009] Plan the flight path and endurance strategy of the drone through the inspection planning model. According to the real-time weather, obstacle information and drone system working status information data, dynamically optimize the flight path and endurance strategy of the drone through the inspection planning model.

[0010] The UAV system performs inspection tasks according to the dynamically optimized flight path and endurance strategy. The image acquisition device carried by the UAV collects the relevant data of the photovoltaic power station in real time during the inspection process, and transmits the relevant data of the photovoltaic power station and the UAV system collected during the inspection process in real time.

[0011] The data analysis and evaluation model is used to analyze and evaluate the real-time data transmitted back from the photovoltaic power station and the data related to the drone system, and corresponding operation and maintenance measures are formulated based on the evaluation results.

[0012] Optionally, the specific process of constructing the inspection planning model based on the inspection management database is:

[0013] Integrate and preprocess the historical data related to inspection planning in the inspection management data and obtain the initial planning data set;

[0014] Based on the actual situation of the photovoltaic power plant and the needs of drone inspections, the key features that affect the inspection planning in the initial planning data set are identified;

[0015] Define decision variables based on key features and establish an initial inspection planning model;

[0016] Based on the initial planning data set, the initial inspection planning model is optimized and trained to obtain the inspection planning model.

[0017] Optionally, the objective function of the initial inspection planning model aims to minimize the weighted sum of total flight time and total energy consumption. The expression of the objective function is shown in the following formula (1):

[0018]

[0019] Among them, W t is the weight of the flight time, W e is the weight of flight energy consumption, n is the total number of inspection points, i and j are the indexes of inspection points, i,j∈{1,2,…,n}, x ij is a binary variable. If the drone flies from point i to point j, then x ij =1, otherwise x ij =0,d ij is the flight distance from point i to point j, e ij Energy consumption for point-to-point flights;

[0020] Set constraints for the objective function. Set the index of the inspection starting point to 1 and the index of the end point to n. Each inspection point is inspected only once. For each inspection point that is not the starting point or the end point, the following equation (2) is satisfied:

[0021]

[0022] The constraints of the inspection start point and inspection end point are shown in the following equations (3) and (4) respectively:

[0023]

[0024] The time window constraint of the UAV is shown in equation (5):

[0025]

[0026] Among them, t i is the time when the UAV arrives at inspection point i, t jis the time when the UAV arrives at the inspection point j, v is the flight speed of the UAV, TW i is the time window of inspection point i, TW j is the time window of inspection point j;

[0027] The maximum flight time and energy consumption constraints are shown in equations (6) and (7) respectively:

[0028]

[0029] Among them, T max is the maximum time limit of a single flight of the drone, E max It is the maximum energy consumption limit of a single flight of the drone.

[0030] Optionally, the specific process of dynamically optimizing the flight path and endurance strategy of the drone through the inspection planning model according to the real-time weather, obstacle information and drone system working status information data is as follows:

[0031] The sensor system installed on the UAV system collects weather, obstacles and UAV system working status information data in real time to obtain a real-time data set, which is pre-processed and fused with historical data in the inspection management database to obtain a comprehensive data set;

[0032] Based on the comprehensive data set, the inspection planning model is used to re-evaluate the flight distance and energy consumption, and the inspection planning model is optimized according to the evaluation results;

[0033] According to the remaining power of the UAV and the energy consumption forecast of the flight path, a corresponding endurance strategy is formulated. If the remaining power is not enough to support the UAV to complete the remaining inspection tasks, a power supply point is set in the flight path or the flight speed is adjusted to reduce energy consumption. If it is predicted that the UAV’s power will be exhausted and power cannot be supplied in time, an emergency strategy is triggered to control the UAV to return to the starting point or land in a safe area.

[0034] Optionally, the specific process of the UAV system performing the inspection task according to the dynamically optimized flight path and endurance strategy is as follows:

[0035] The UAV system performs inspection tasks according to the dynamically optimized flight path and endurance strategy, controls several UAVs to perform inspection tasks, and maintains coordination and cooperation between UAVs through real-time communication;

[0036] Monitor the flight status and inspection progress of each drone in real time, and adjust the inspection path and speed of the drones in the inspection mission according to the flight status and inspection progress of the drones. If the inspection progress is significantly delayed, dispatch spare drones to coordinate.

[0037] Optionally, the specific process of constructing the data analysis and evaluation model based on the inspection management database is:

[0038] Preprocess the relevant data of photovoltaic power plants and drone systems in the inspection management database and obtain analysis and evaluation data sets;

[0039] Extract features from the analysis and evaluation data set and build an initial data analysis and evaluation model;

[0040] The initial data analysis and evaluation model is trained by analyzing the evaluation data set to obtain the data analysis and evaluation model, and the data analysis and evaluation model is iteratively optimized.

[0041] Optionally, the specific process of preprocessing the relevant data of the photovoltaic power station and the drone system in the inspection management database to obtain the analysis and evaluation data set is:

[0042] Preprocess the photovoltaic power station image data collected by drones, and clean, convert and normalize the non-image data;

[0043] The image data is extracted using a pre-trained deep learning model, and non-image data is directly used as feature input;

[0044] The features extracted from the image data and the features of the non-image data are fused to obtain an analysis and evaluation data set.

[0045] Optionally, the expression of the data analysis and evaluation model is as shown in the following formula (8):

[0046]

[0047] Among them, suppose there are m base learners, is the model's predicted output for input x, F(x) is the function of the entire data analysis and evaluation model, T k (x) is the prediction output of the kth base learner for input x. Formula (8) represents the simple average of the prediction outputs of m base learners to obtain the final prediction result

[0048] Optionally, the specific process of formulating corresponding operation and maintenance measures according to the evaluation results is:

[0049] Prioritize the problems found based on the evaluation results, and develop specific operation and maintenance plans for each priority problem;

[0050] According to the operation and maintenance plan, rationally allocate and dispatch resources, perform various operation and maintenance tasks and establish a supervision mechanism;

[0051] After the operation and maintenance task is completed, the data of the photovoltaic power station and the drone system are collected again, compared with the data before the operation and maintenance, and the operation and maintenance plan is evaluated;

[0052] Based on the evaluation results of the operation and maintenance plan, the existing operation and maintenance processes, methods and tools are continuously improved and optimized.

[0053] The present invention also provides a system for implementing the photovoltaic inspection management method based on a drone, comprising:

[0054] Database construction module, used to collect and organize historical data related to photovoltaic power plants and drone systems, and build an inspection management database;

[0055] Model building module, used to build inspection planning model and data analysis and evaluation model;

[0056] The flight planning module is used to dynamically optimize the flight path and endurance strategy of the drone based on real-time weather, obstacle information and the working status of the drone system;

[0057] The mission execution module is used to control the UAV system to perform inspection tasks according to the dynamically optimized flight path and endurance strategy;

[0058] A data analysis and evaluation module is used to receive real-time data from the drone and analyze and evaluate the data using a data analysis and evaluation model;

[0059] The operation and maintenance decision-making module is used to formulate specific operation and maintenance measures based on the evaluation results of the data analysis and evaluation module, and to rationally allocate and schedule resources to perform operation and maintenance tasks, collect new data for effect evaluation, and continuously improve and optimize the operation and maintenance processes, methods and tools based on the effect evaluation results.

[0060] The technical solution of the embodiment of the present invention has at least the following advantages and beneficial effects:

[0061] Improve inspection efficiency and accuracy: PV inspections by drones can quickly cover a large area of ​​PV power stations, effectively reducing the time and cost of manual inspections. At the same time, the high-definition image acquisition device carried by the drone can collect status information of key parts such as PV modules, brackets, cables, etc. in real time and accurately, improving the accuracy and comprehensiveness of inspections.

[0062] Intelligence and automation: The inspection planning model and data analysis and evaluation model constructed by the present invention can intelligently plan the flight path and endurance strategy of the drone based on historical data and real-time information, and dynamically optimize according to actual conditions, which greatly reduces the errors caused by human factors and improves the efficiency and reliability of inspection work;

[0063] Real-time data analysis and evaluation: The data analysis and evaluation model can quickly process and analyze the real-time data sent back, and timely discover potential problems of the photovoltaic power station, such as component damage, dirt accumulation, cable aging, etc., to provide strong decision-making support for operation and maintenance personnel, help take measures in advance, avoid the expansion of faults, and ensure the safe and stable operation of the photovoltaic power station;

[0064] Optimize resource allocation: Based on the data analysis and evaluation results, the operation and maintenance decision-making module can reasonably allocate and dispatch resources and formulate specific operation and maintenance plans, which can not only improve resource utilization efficiency but also reduce operation and maintenance costs. At the same time, through continuous effect evaluation and optimization, the operation and maintenance processes, methods and tools are continuously improved to further improve the operation and maintenance management level;

[0065] Enhanced emergency response capability: The present invention takes into account the power supply point and emergency strategy in the drone endurance strategy, ensuring that when the drone is low on power or encounters an emergency, it can respond quickly, ensuring the safety of the drone and inspection tasks, which is of great significance for improving the reliability and stability of the photovoltaic power station;

[0066] Collaborative operation and multi-machine dispatching: By controlling multiple drones to work collaboratively, the present invention can further improve the inspection efficiency and dispatch spare drones to provide support in a timely manner when the inspection progress lags behind, making the entire inspection system more flexible and efficient;

[0067] Data-driven decision-making: The entire inspection management method is centered on data. By collecting, processing, analyzing and evaluating massive amounts of data, it provides a scientific basis for operation and maintenance decisions, helping to achieve refined, intelligent and efficient operation and maintenance management of photovoltaic power stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 This is a flow chart of a photovoltaic inspection management method based on a drone in Embodiment 1 of the present invention;

[0069] Figure 2 This is a schematic diagram of the structure of a photovoltaic inspection management system based on a drone in Example 3 of the present invention. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0071] Example 1

[0072] Reference Figure 1, a photovoltaic inspection management method based on drones, comprising the steps of:

[0073] Step 1: Build an inspection management database based on historical data related to photovoltaic power plants and drone systems;

[0074] In this embodiment, the historical data related to the photovoltaic power station includes the geographical location, layout diagram, photovoltaic panel model, installation year, historical power generation, fault records, etc. of the photovoltaic power station; the historical data related to the drone system includes the model of the drone, flight speed, maximum flight time, battery capacity, historical flight records (including flight path, flight time, energy consumption, etc.), sensor data (such as GPS positioning, altimeter, meteorological sensor, etc.); the database also includes historical inspection data, flight paths, inspection points, inspection times, inspection results (such as damage to photovoltaic panels, obstructions, stains, etc.), weather conditions during the inspection, drone system status, etc.

[0075] In this embodiment, the collected data is cleaned and preprocessed. Duplicate, missing or obviously erroneous data records are deleted; the data from different sources are unified in format, such as the timestamp is unified to UTC time, and the geographic location data is converted to a unified coordinate system; the numerical data is standardized, such as the battery power, flight speed, etc. are converted into relative values ​​or percentages to facilitate subsequent analysis.

[0076] In this embodiment, the data is integrated. The database table structure is designed, including the photovoltaic power station information table, the drone system information table, the inspection record table, the weather data table, etc. The cleaned and pre-processed data is imported into the database to ensure the integrity and consistency of the data; indexes are established for key fields (such as inspection point ID, drone ID, timestamp, etc.) to improve query efficiency.

[0077] In this embodiment, the data is verified and tested. Verify whether the data in the database is complete, without omissions or errors; check whether the associated data between different tables are consistent, such as whether the drone ID in the inspection record matches the ID in the drone system information table; perform performance testing on the database to ensure that it can remain stable and efficient during high-concurrency queries.

[0078] In this embodiment, the database is maintained and updated. The database is backed up regularly to prevent data loss or damage; as the photovoltaic power station and the drone system operate, new data is continuously collected and updated to the database; a database monitoring and log system is established to record the operating status and abnormal information of the database to facilitate problem troubleshooting and performance optimization.

[0079] Step 2: construct an inspection planning model and a data analysis and evaluation model based on the inspection management database;

[0080] In this embodiment, the specific process of constructing the inspection planning model is as follows:

[0081] The historical data related to the inspection plan in the inspection management data are integrated and preprocessed to obtain an initial planning data set; in this embodiment, historical data related to the inspection plan are extracted from the inspection management database, including but not limited to the location of the inspection point, historical flight path, flight time, flight energy consumption, weather conditions, drone system status, etc.; these data are cleaned to remove invalid, abnormal or duplicate records; the data are preprocessed, such as standardizing numerical data such as flight time and energy consumption, and converting text data into numerical encoding or unique hot encoding; the processed data are integrated into an initial planning data set, which contains key information such as the distance between inspection points and energy consumption.

[0082] Based on the actual situation of the photovoltaic power station and the needs of drone inspections, the key features that affect the inspection planning in the initial planning data set are identified; in this embodiment, based on the layout, terrain, obstacle distribution of the photovoltaic power station and the performance characteristics of the drone (such as flight speed, endurance, load capacity, etc.), the key features that affect the inspection planning are identified; these features may include the location of the inspection point, height difference, wind speed, temperature, light intensity, flight speed of the drone, energy consumption rate, etc.

[0083] Based on the key features, decision variables are defined and an initial inspection planning model is established. In this embodiment, the objective function of the initial inspection planning model aims to minimize the weighted sum of the total flight time and the total energy consumption. The expression of the objective function is shown in the following formula (1):

[0084]

[0085] Among them, W t is the weight of the flight time, W e is the weight of flight energy consumption, n is the total number of inspection points, i and j are the indexes of inspection points, i,j∈{1,2,…,n}, x ij is a binary variable. If the drone flies from point i to point j, then x ij =1, otherwise x ij =0,d ij is the flight distance from point i to point j, e ij Energy consumption for point-to-point flights;

[0086] Set constraints for the objective function. Set the index of the inspection starting point to 1 and the index of the end point to n. Each inspection point is inspected only once. For each inspection point that is not the starting point or the end point, the following equation (2) is satisfied:

[0087]

[0088] The constraints of the inspection start point and inspection end point are shown in the following equations (3) and (4) respectively:

[0089]

[0090] The time window constraint of the UAV is shown in equation (5):

[0091]

[0092] Among them, t i is the time when the UAV arrives at inspection point i, t j is the time when the UAV arrives at the inspection point j, v is the flight speed of the UAV, TW i is the time window of inspection point i, TW j is the time window of inspection point j;

[0093] The maximum flight time and energy consumption constraints are shown in equations (6) and (7) respectively:

[0094]

[0095] Among them, T max is the maximum time limit of a single flight of the drone, E max The maximum energy consumption limit for a single flight of the drone;

[0096] Based on the initial planning data set, the initial inspection planning model is optimized and trained to obtain the inspection planning model.

[0097] In this embodiment, the specific process of building a data analysis and evaluation model is as follows:

[0098] Preprocess the relevant data of photovoltaic power plants and drone systems in the inspection management database and obtain analysis and evaluation data sets;

[0099] Preprocess the photovoltaic power station image data collected by the drone. The image preprocessing method can be to perform denoising, contrast enhancement, image segmentation and other processing on the photovoltaic power station images collected by the drone to improve the image quality and facilitate subsequent feature extraction;

[0100] Clean, convert, and normalize non-image data to remove invalid, abnormal, or missing data, convert non-image data into a format suitable for model processing, and normalize it to ensure that data of different dimensions are at the same numerical level to avoid deviation during model training;

[0101] The pre-trained deep learning model is used to extract features from image data. These image features can reflect key information in the image, such as damage, occlusion, and stains on photovoltaic panels.

[0102] Non-image data is directly used as feature input. These non-image features include environmental parameters such as temperature, humidity, light intensity, and state parameters of the UAV system, such as battery power, flight speed, altitude, etc.

[0103] The features extracted from the image data are fused with the features of the non-image data to obtain an analysis and evaluation data set;

[0104] Extract features from the analysis and evaluation data set and build an initial data analysis and evaluation model. The initial data analysis and evaluation model can use a variety of machine learning or deep learning algorithms, such as random forest, gradient boosted tree (GBDT), neural network, etc.

[0105] The data analysis and evaluation model is obtained by training the initial data analysis and evaluation model with the analysis and evaluation data set. The model parameters can be continuously adjusted using the gradient descent method. The initial model is trained with the analysis and evaluation data set to improve the prediction accuracy and generalization ability of the model.

[0106] During the training process, multiple iterative optimizations are performed based on the performance of the model, including adjusting the model structure, optimizing algorithm parameters, adding regularization terms, etc., to iteratively optimize the data analysis and evaluation model.

[0107] The expression of the data analysis evaluation model is shown in the following formula (8):

[0108]

[0109] Among them, suppose there are m base learners, is the predicted output of the model for the input x, representing the evaluation result of the status or performance of the photovoltaic power station, such as the degree of damage to the photovoltaic panels, power generation efficiency, etc.; F(x) is the function of the entire data analysis and evaluation model, T k (x) is the prediction output of the kth base learner for the input x. Each base learner can be an independent machine learning model, such as a decision tree, a neural network, a support vector machine, etc. In this embodiment, these base learners can respectively learn different aspects or features of the photovoltaic power station; Formula (8) represents the simple average of the prediction outputs of m base learners to obtain the final prediction result Reduce the variance of the error and improve the stability of the model.

[0110] Step 3: Plan the flight path and endurance strategy of the drone through the inspection planning model. According to the real-time weather, obstacle information and drone system working status information data, the flight path and endurance strategy of the drone are dynamically optimized through the inspection planning model.

[0111] In this embodiment, a sensor system installed on the drone system collects weather, obstacles and drone system working status information data in real time to obtain a real-time data set, which is pre-processed and fused with historical data in the inspection management database to obtain a comprehensive data set;

[0112] Based on the comprehensive data set, the inspection planning model is used to re-evaluate the flight distance and energy consumption, and the inspection planning model is optimized according to the evaluation results;

[0113] According to the remaining power of the UAV and the energy consumption forecast of the flight path, a corresponding endurance strategy is formulated. If the remaining power is not enough to support the UAV to complete the remaining inspection tasks, a power supply point is set in the flight path or the flight speed is adjusted to reduce energy consumption. If it is predicted that the UAV’s power will be exhausted and power cannot be supplied in time, an emergency strategy is triggered to control the UAV to return to the starting point or land in a safe area.

[0114] Step 4: The UAV system performs the inspection task according to the dynamically optimized flight path and endurance strategy. The image acquisition device carried by the UAV collects the relevant data of the photovoltaic power station during the inspection in real time, and transmits the relevant data of the photovoltaic power station and the UAV system collected during the inspection in real time.

[0115] In this embodiment, the drone system performs inspection tasks according to the dynamically optimized flight path and endurance strategy, controls several drones to perform inspection tasks, and maintains the coordination and cooperation between the drones to perform inspection tasks through real-time communication;

[0116] Monitor the flight status and inspection progress of each drone in real time, and adjust the inspection path and speed of the drones in the inspection mission according to the flight status and inspection progress of the drones. If the inspection progress is significantly delayed, dispatch spare drones to coordinate.

[0117] Step 5: Analyze and evaluate the relevant data of the photovoltaic power station and the relevant data of the drone system transmitted back in real time through the data analysis and evaluation model, and formulate corresponding operation and maintenance measures based on the evaluation results;

[0118] In this embodiment, the specific process of formulating corresponding operation and maintenance measures according to the evaluation results is as follows:

[0119] Prioritize the problems found based on the evaluation results, and develop specific operation and maintenance plans for each priority problem. For example, if a photovoltaic panel is damaged, a maintenance team can be arranged to go to the site to replace or repair it; if a drone is running low on battery, the flight path can be adjusted to reduce energy consumption or a power supply point can be set;

[0120] According to the operation and maintenance plan, rationally allocate and dispatch resources, including human, material and financial resources, etc., perform various operation and maintenance tasks and establish a supervision mechanism;

[0121] After the operation and maintenance task is completed, the data of the photovoltaic power station and the drone system are collected again, compared with the data before the operation and maintenance, and the operation and maintenance plan is evaluated;

[0122] Based on the evaluation results of the operation and maintenance plan, the existing operation and maintenance processes, methods and tools are continuously improved and optimized; for example, the algorithm of the data analysis and evaluation model is optimized, the stability and reliability of the drone system is improved, and the operation and maintenance plan is improved. Through continuous improvement and optimization, the efficiency and accuracy of photovoltaic inspections are improved and the operation and maintenance costs are reduced.

[0123] Example 2

[0124] Based on Example 1, in this embodiment, based on the initial planning data set, the specific process of optimizing and training the initial inspection planning model and obtaining the inspection planning model is as follows:

[0125] In this embodiment, the initial inspection planning model is optimized and trained by the particle swarm optimization algorithm. First, the particle swarm size is set to N, and each particle represents a possible inspection path solution, that is, a set of x ij (binary variables), randomly generate the initial position of each particle, that is, randomly generate a set of x that satisfy the constraints ij , assign an initial velocity to each particle. The velocity can be a random decimal value, indicating the direction and amplitude of the position change. During initialization, the fitness value (i.e., the objective function value) of each particle is set to the global optimal and individual optimal initial values.

[0126] For each iteration, perform the following steps:

[0127] Calculate the fitness,according to the objective function, as shown in formula (1), calculate the fitness value of each particle;

[0128] Update the individual optimum. If the fitness value of the current particle is better than the fitness value of its individual optimum, then update the individual optimum.

[0129] Update the global optimum, and select the one with the best fitness value among all the individual optima understood as the global optimum;

[0130] Update the speed and position. Update the speed and position of each particle according to the following equations (9) and (10). Equations (9) and (10) are as follows:

[0131]

[0132] in, are the i-th particle in the d-th dimension (i.e., the d-th x ij ) is the speed and position of the tth iteration on the particle; ω is the inertia weight, which is used to control the degree to which the particle maintains its original speed; c1 and c2 are learning factors, which represent the degree to which the particle learns from the individual optimum and the global optimum, respectively; r1 and r2 are random numbers between [0,1]; is the optimal individual position of the i-th particle in the d-th dimension; is the global optimal position.

[0133] In this embodiment, since it is a binary variable, the speed needs to be converted into a probability for the update of the decision, and the conversion is performed through the Sigma ID function to ensure that the updated position and speed are within the allowed range. For binary variables, the speed is converted to 0 or 1 through a threshold (such as 0.5), and the position update expression after conversion is shown in the following formula (11):

[0134]

[0135] Among them, tHreshold is usually set to 0.5; when the preset maximum number of iterations is reached or other termination conditions are met, such as when the fitness value no longer changes significantly, the iteration is stopped and the global optimal solution is output as the final inspection planning model, that is, the optimal inspection path and endurance strategy.

[0136] Example 3

[0137] Reference Figure 2 Based on Example 1, in this embodiment, a system for implementing a photovoltaic inspection management method based on a drone is provided, including:

[0138] The database construction module is used to collect and organize historical data related to photovoltaic power stations and drone systems, including photovoltaic power station operation data, drone flight records, weather data, fault maintenance records and other related data, to build an inspection management database; obtain data from photovoltaic power station management systems, drone control systems, meteorological stations and other data sources through data interfaces, and clean, integrate and store data to build an inspection management database.

[0139] Model building module, used to build inspection planning model and data analysis and evaluation model;

[0140] The flight planning module is used to dynamically optimize the flight path and endurance strategy of the drone based on real-time weather, obstacle information and the working status of the drone system;

[0141] The mission execution module is used to control the UAV system to perform inspection tasks according to the dynamically optimized flight path and endurance strategy;

[0142] A data analysis and evaluation module is used to receive real-time data from the drone and analyze and evaluate the data using a data analysis and evaluation model;

[0143] The operation and maintenance decision-making module is used to formulate specific operation and maintenance measures based on the evaluation results of the data analysis and evaluation module, and to rationally allocate and schedule resources to perform operation and maintenance tasks, collect new data for effect evaluation, and continuously improve and optimize the operation and maintenance processes, methods and tools based on the effect evaluation results.

[0144] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A photovoltaic inspection management method based on drones, characterized in that: Includes steps: Build an inspection management database based on historical data related to photovoltaic power plants and drone systems; Based on the inspection management database, the inspection planning model and data analysis and evaluation model are constructed respectively; Plan the flight path and endurance strategy of the drone through the inspection planning model. According to the real-time weather, obstacle information and drone system working status information data, dynamically optimize the flight path and endurance strategy of the drone through the inspection planning model. The UAV system performs inspection tasks according to the dynamically optimized flight path and endurance strategy. The image acquisition device carried by the UAV collects the relevant data of the photovoltaic power station in real time during the inspection process, and transmits the relevant data of the photovoltaic power station and the UAV system collected during the inspection process in real time. The data analysis and evaluation model is used to analyze and evaluate the real-time data transmitted back from the photovoltaic power station and the data related to the drone system, and corresponding operation and maintenance measures are formulated based on the evaluation results.

2. The photovoltaic inspection management method based on drone according to claim 1, characterized in that: The specific process of constructing the inspection planning model based on the inspection management database is as follows: Integrate and preprocess the historical data related to inspection planning in the inspection management data and obtain the initial planning data set; Based on the actual situation of the photovoltaic power station and the demand for drone inspection, the key features that affect the inspection planning in the initial planning data set are identified; Define decision variables based on key features and establish an initial inspection planning model; Based on the initial planning data set, the initial inspection planning model is optimized and trained to obtain the inspection planning model.

3. The photovoltaic inspection management method based on drone according to claim 2, characterized in that: The objective function of the initial inspection planning model aims to minimize the weighted sum of total flight time and total energy consumption. The expression of the objective function is shown in the following formula (1): Among them, W t is the weight of the flight time, W e is the weight of flight energy consumption, n is the total number of inspection points, i and j are the indexes of inspection points, i,j∈{1,2,…,n}, x ij is a binary variable. If the drone flies from point i to point j, then x ij =1, otherwise x ij =0,d ij is the flight distance from point i to point j, e ij Energy consumption for point-to-point flights; Set constraints for the objective function. Set the index of the inspection starting point to 1 and the index of the end point to n. Each inspection point is inspected only once. For each inspection point that is not the starting point or the end point, the following equation (2) is satisfied: The constraints of the inspection start point and inspection end point are shown in the following equations (3) and (4) respectively: The time window constraint of the UAV is shown in equation (5): Among them, t i is the time when the UAV arrives at inspection point i, t j is the time when the UAV arrives at the inspection point j, v is the flight speed of the UAV, TW i is the time window of inspection point i, TW j is the time window of inspection point j; The maximum flight time and energy consumption constraints are shown in equations (6) and (7) respectively: Among them, T max is the maximum time limit of a single flight of the drone, E max It is the maximum energy consumption limit of a single flight of the drone.

4. The photovoltaic inspection management method based on drone according to claim 1, characterized in that: The specific process of dynamically optimizing the flight path and endurance strategy of the drone through the inspection planning model based on real-time weather, obstacle information and drone system working status information data is as follows: The sensor system installed on the UAV system collects weather, obstacles and UAV system working status information data in real time to obtain a real-time data set, which is pre-processed and fused with historical data in the inspection management database to obtain a comprehensive data set; Based on the comprehensive data set, the inspection planning model is used to re-evaluate the flight distance and energy consumption, and the inspection planning model is optimized according to the evaluation results; According to the remaining power of the UAV and the energy consumption forecast of the flight path, a corresponding endurance strategy is formulated. If the remaining power is not enough to support the UAV to complete the remaining inspection tasks, a power supply point is set in the flight path or the flight speed is adjusted to reduce energy consumption. If it is predicted that the UAV’s power will be exhausted and power cannot be supplied in time, an emergency strategy is triggered to control the UAV to return to the starting point or land in a safe area.

5. The photovoltaic inspection management method based on drone according to claim 1, characterized in that: The specific process of the UAV system performing the inspection task according to the dynamically optimized flight path and endurance strategy is as follows: The UAV system performs inspection tasks according to the dynamically optimized flight path and endurance strategy, controls several UAVs to perform inspection tasks, and maintains coordination and cooperation between UAVs through real-time communication; Monitor the flight status and inspection progress of each drone in real time, and adjust the inspection path and speed of the drones in the inspection mission according to the flight status and inspection progress of the drones. If the inspection progress is significantly delayed, dispatch spare drones to coordinate.

6. The photovoltaic inspection management method based on drone according to claim 1, characterized in that: The specific process of constructing the data analysis and evaluation model based on the inspection management database is as follows: Preprocess the relevant data of photovoltaic power plants and drone systems in the inspection management database and obtain analysis and evaluation data sets; Extract features from the analysis and evaluation data set and build an initial data analysis and evaluation model; The initial data analysis and evaluation model is trained by analyzing the evaluation data set to obtain the data analysis and evaluation model, and the data analysis and evaluation model is iteratively optimized.

7. The photovoltaic inspection management method based on drone according to claim 6, characterized in that: The specific process of preprocessing the relevant data of the photovoltaic power station and the drone system in the inspection management database and obtaining the analysis and evaluation data set is as follows: Preprocess the photovoltaic power station image data collected by drones, and clean, convert and normalize the non-image data; The image data is extracted using a pre-trained deep learning model, and non-image data is directly used as feature input; The features extracted from the image data and the features of the non-image data are fused to obtain an analysis and evaluation data set.

8. The photovoltaic inspection management method based on drone according to claim 6, characterized in that: The expression of the data analysis and evaluation model is shown in the following formula (8): Among them, suppose there are m base learners, is the model's predicted output for input x, F(x) is the function of the entire data analysis and evaluation model, T k (x) is the prediction output of the kth base learner for input x. Formula (8) represents the simple average of the prediction outputs of m base learners to obtain the final prediction result 9. The photovoltaic inspection management method based on drone according to claim 1, characterized in that: The specific process of formulating corresponding operation and maintenance measures based on the evaluation results is as follows: Prioritize the problems found based on the evaluation results, and develop specific operation and maintenance plans for each priority problem; According to the operation and maintenance plan, rationally allocate and dispatch resources, perform various operation and maintenance tasks and establish a supervision mechanism; After the operation and maintenance task is completed, the data of the photovoltaic power station and the drone system are collected again, compared with the data before the operation and maintenance, and the operation and maintenance plan is evaluated; Based on the evaluation results of the operation and maintenance plan, the existing operation and maintenance processes, methods and tools are continuously improved and optimized.

10. A system for implementing the photovoltaic inspection management method based on drones as described in any one of claims 1 to 9, characterized in that: include: Database construction module, used to collect and organize historical data related to photovoltaic power plants and drone systems, and build an inspection management database; Model building module, used to build inspection planning model and data analysis and evaluation model; The flight planning module is used to dynamically optimize the flight path and endurance strategy of the drone based on real-time weather, obstacle information and the working status of the drone system; The mission execution module is used to control the UAV system to perform inspection tasks according to the dynamically optimized flight path and endurance strategy; A data analysis and evaluation module is used to receive real-time data from the drone and analyze and evaluate the data using a data analysis and evaluation model; The operation and maintenance decision-making module is used to formulate specific operation and maintenance measures based on the evaluation results of the data analysis and evaluation module, and to rationally allocate and schedule resources to perform operation and maintenance tasks, collect new data for effect evaluation, and continuously improve and optimize the operation and maintenance processes, methods and tools based on the effect evaluation results.

Images