Multi-mode cooperative photovoltaic power station inspection method and system

Abstract

The invention relates to the technical field of photovoltaic power station inspection, and discloses a multi-mode cooperative photovoltaic power station inspection method and system, and the method comprises the steps: obtaining the position information, spectral reflection data and initial geofence of a power station; performing aging analysis according to the spectral reflection data to obtain material aging characteristics; inputting the material aging characteristics into a pre-trained aging evaluation model, and outputting an aging evaluation result; performing spatial clustering according to the power station position information to obtain power station clustering data; carrying out iterative optimization according to the power station clustering data and the initial geo-fence to obtain an optimized geo-fence; according to the optimized geofence and the aging evaluation result, routing inspection path planning is carried out, an optimal routing inspection path is obtained, and routing inspection is carried out on the photovoltaic power station according to the optimal routing inspection path. The method has the following effect that the inspection efficiency of the photovoltaic power station can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power station inspection, and in particular to a multi-modal collaborative photovoltaic power station inspection method and system. Background Art

[0002] As an important part of sustainable energy solutions, the efficient operation of photovoltaic power stations is crucial to promoting the development of green energy. However, in the actual inspection process, photovoltaic power stations face many challenges. Since photovoltaic modules are installed in a wide area, manual inspection is not only time-consuming but also labor-intensive, especially in areas with complex terrain or harsh climatic conditions. In addition, traditional inspection methods make it difficult to achieve a comprehensive inspection of all photovoltaic panels, resulting in certain problems not being discovered in time, such as hidden cracks and hot spot effects. If these problems are not handled in time, they will directly affect the power generation efficiency and service life of the power station. Therefore, how to effectively improve the inspection efficiency and ensure the quality of inspection has become a key issue in the operation and maintenance of photovoltaic power stations.

[0003] In one prior art, in order to improve the efficiency and accuracy of photovoltaic power station inspections, a widely used technical means is to use drones equipped with multi-sensor systems for automated inspections. First, the drone is equipped with a high-resolution RGB camera, an infrared thermal imager, and a laser radar (LiDAR) to capture visible light images, thermal images, and three-dimensional spatial information, respectively. Before the inspection mission begins, a detailed flight plan is formulated based on the specific layout and environmental characteristics of the photovoltaic power station, and automatic obstacle avoidance parameters are set to ensure that the drone can complete the mission safely and effectively. During the flight, the drone autonomously navigates along the preset path, collects data in real time, and synchronizes the data to the ground station through wireless transmission technology. Afterwards, the collected data is analyzed using image processing algorithms, for example, by comparing thermal images at different time points to identify abnormal heating areas, or by applying computer vision technology to detect physical damage from RGB images. These analysis results can quickly locate the fault location and provide accurate guidance for subsequent maintenance work.

[0004] However, although the multi-sensor inspection system based on drones has improved the efficiency and accuracy of PV power station inspections, its efficiency is still limited when facing large-scale power stations. Specifically, when the PV power station covers a very large area, it becomes extremely time-consuming to use drones to inspect each area one by one, resulting in an extension of the overall inspection cycle. In fact, the photovoltaic panels in many areas are in good condition, and it is easy to determine that there is no obvious damage or abnormality through conventional high-definition pictures, without the need for further detailed inspections with the help of thermal imaging or other advanced sensors. In this case, the full-coverage inspection method of drones not only wastes time, but also increases unnecessary workload, resulting in low efficiency of PV power station inspections. Summary of the invention

[0005] The present invention provides a multi-modal collaborative photovoltaic power station inspection method and system to improve the efficiency of photovoltaic power station inspection.

[0006] In a first aspect, in order to solve the above technical problems, the present invention provides a multi-modal collaborative photovoltaic power station inspection method, comprising: Obtain power station location information, spectral reflectance data, and initial geo-fence; Performing aging analysis based on the spectral reflectance data to obtain material aging characteristics; Inputting the material aging characteristics into a pre-trained aging assessment model and outputting an aging assessment result; Performing spatial clustering according to the power station location information to obtain power station clustering data; Iteratively optimizing the power station clustering data and the initial geo-fence to obtain an optimized geo-fence; Inspection paths are planned according to the optimized geographic fence and the aging assessment result to obtain an optimal inspection path, and the photovoltaic power station is inspected according to the optimal inspection path.

[0007] In an optional implementation, performing aging analysis according to the spectral reflectance data to obtain material aging characteristics includes: Performing a linear regression analysis based on the spectral reflectance data and a preset first wavelength range to obtain a spectral slope; Performing Gaussian fitting based on the spectral reflection data and a preset second wavelength range to obtain a wavelength reflection function; Performing a second-order derivative calculation based on the wavelength reflection function to obtain a reflection second-order derivative; The wavelength corresponding to the maximum value of the second-order derivative of the reflection is taken as the reflection peak position; constructing a spectral matrix according to the spectral reflectance data; Performing principal component analysis on the spectral matrix to obtain a dimension-reduced spectral matrix; The pollution index is calculated according to the following formula: in, represents the pollution index, Indicates the dimension number of the reduced spectral matrix, represents the total number of dimensions of the reduced spectral matrix, express The weight coefficient of the dimension, Represents the reduced spectral matrix The value of the dimension; The material aging characteristics include the spectral slope, the reflection peak position and the pollution index.

[0008] In an optional implementation, the training process of the aging assessment model includes: An aging assessment model is constructed based on historical aging characteristics and historical aging results, and the model is trained. When the number of training times reaches a set upper limit or the loss function of the detection model meets the conditions, the training is considered complete, and the trained model is obtained; The material aging characteristics are input into the trained model to obtain an aging assessment result.

[0009] In an optional implementation, performing spatial clustering according to the power station location information to obtain power station clustering data includes: Performing coordinate system conversion on the power station location information to obtain plane location data; Get the neighborhood radius and minimum number of core points; traversing the plane position data; Marking points within the neighborhood radius whose number of coordinate points is greater than the minimum number of core points as core points; Marking coordinate points in the neighborhood of the core point that do not meet the core standard as boundary points, and merging the boundary points into the core point cluster of the core point; The coordinate points that are not incorporated into any core point cluster are marked as noise points; The power station clustering data includes the distribution of the core points, the boundary points and the noise points.

[0010] In an optional implementation, the iterative optimization is performed according to the power station clustering data and the initial geo-fence to obtain an optimized geo-fence, including: generating an initial convex hull fence according to the power station clustering data; Performing spatial topological analysis based on the initial geographic fence and the power station clustering data to obtain an abnormal fence; Optimizing the abnormal fence according to a preset fence optimization rule to obtain an updated geographic fence; The updated geo-fence and the initial convex hull fence are merged to obtain an optimized geo-fence.

[0011] In an optional implementation, performing spatial topology analysis based on the initial geographic fence and the power station clustering data to obtain an abnormal fence includes: When there is a core point outside the initial geo-fence, it is determined to be an out-of-bounds component fence; When there is no core point in the initial geographic fence, it is determined to be an empty area fence; Calculating power station density according to the initial geographic fence and the power station clustering data to obtain power station density; When the peak value difference of the power station density within the initial geographic fence is greater than a preset density threshold, it is determined to be a density mutation area fence.

[0012] In an optional implementation, the performing inspection path planning according to the optimized geographic fence and the aging assessment result to obtain an optimal inspection path, and inspecting the photovoltaic power station according to the optimal inspection path includes: Divide the inspection area according to the optimized geo-fence; marking high-risk components based on the aging assessment results and the optimized geo-fencing; Initialize a basic inspection path that can traverse all the inspection areas; generating key inspection points according to the high-risk components; Inserting the key detection points into the basic inspection path to obtain an updated inspection path; The updated inspection path is smoothed to obtain an optimal inspection path.

[0013] In a second aspect, the present invention provides a multi-modal collaborative photovoltaic power station inspection system, comprising: A data acquisition module, used to obtain power station location information, spectral reflectance data and initial geo-fence; An aging analysis module, used to perform aging analysis based on the spectral reflectance data to obtain material aging characteristics; An aging assessment module, used for inputting the material aging characteristics into a pre-trained aging assessment model and outputting an aging assessment result; A power station clustering module, used for performing spatial clustering according to the power station location information to obtain power station clustering data; A fence optimization module, used for iteratively optimizing according to the power station clustering data and the initial geo-fence to obtain an optimized geo-fence; A path planning module is used to plan an inspection path according to the optimized geographic fence and the aging assessment result, obtain an optimal inspection path, and inspect the photovoltaic power station according to the optimal inspection path.

[0014] In a third aspect, the present invention further provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the multimodal collaborative photovoltaic power station inspection method described in any one of the above is implemented.

[0015] In a fourth aspect, the present invention further provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the multi-modal collaborative photovoltaic power station inspection methods described above.

[0016] Compared with the prior art, the present invention has the following beneficial effects: the present invention discloses a multi-modal collaborative photovoltaic power station inspection method and system, the method comprising obtaining power station location information, spectral reflection data and initial geographic fence; performing aging analysis based on the spectral reflection data to obtain material aging characteristics; inputting the material aging characteristics into a pre-trained aging assessment model to output aging assessment results; performing spatial clustering based on the power station location information to obtain power station clustering data; performing iterative optimization based on the power station clustering data and the initial geographic fence to obtain an optimized geographic fence; performing inspection path planning based on the optimized geographic fence and the aging assessment results to obtain an optimal inspection path, and inspecting the photovoltaic power station according to the optimal inspection path. The present method has the following effects: the present method can improve the inspection efficiency of photovoltaic power stations.

[0017] Specifically, by obtaining power station location information, spectral reflectance data and initial geographic fences, a comprehensive data foundation is provided for subsequent analysis. In particular, aging analysis using spectral reflectance data can accurately identify the aging characteristics of materials. This method is more scientific and accurate than traditional time-based or experience-based assessments and can effectively locate areas that need attention.

[0018] Furthermore, the material aging characteristics are input into the pre-trained aging assessment model to output the aging assessment results. This process not only improves the accuracy of the assessment, but also speeds up the assessment process. Since the model is pre-trained, it means that it has learned patterns and rules from a large amount of historical data, so that it can provide fast and reliable prediction results on new data.

[0019] Furthermore, spatial clustering is performed based on the power station location information to obtain power station cluster data, and iterative optimization is performed in combination with the initial geo-fence to obtain the optimized geo-fence. This approach helps to improve the accuracy and practicality of the geo-fence, allowing inspection activities to cover the target area more efficiently, while reducing unnecessary repeated inspections and improving inspection efficiency.

[0020] Furthermore, the optimal inspection route is planned based on the optimized geo-fencing and aging assessment results. This step fully considers the actual status and distribution of the power station to ensure the pertinence and effectiveness of the inspection work. The inspection route generated by the intelligent algorithm can minimize the inspection time while ensuring that all key points are properly inspected.

[0021] In summary, this invention has shown significant advantages in improving the inspection efficiency of photovoltaic power plants by integrating multiple data sources and applying advanced data analysis methods. It can not only more accurately identify components that need maintenance, but also optimize inspection routes and improve the efficiency of photovoltaic power plant inspections. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic flow chart of a multi-modal collaborative photovoltaic power station inspection method provided by the first embodiment of the present invention; Figure 2 It is a schematic diagram of the structure of a multi-modal collaborative photovoltaic power station inspection system provided by the second embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] Reference Figure 1 The first embodiment of the present invention provides a multi-modal collaborative photovoltaic power station inspection method, comprising the following steps: S11, obtaining power station location information, spectral reflectance data and initial geo-fence; S12, performing aging analysis according to the spectral reflectance data to obtain material aging characteristics; S13, inputting the material aging characteristics into a pre-trained aging assessment model, and outputting an aging assessment result; S14, performing spatial clustering according to the power station location information to obtain power station clustering data; S15, performing iterative optimization according to the power station clustering data and the initial geo-fence to obtain an optimized geo-fence; S16, performing inspection path planning according to the optimized geographic fence and the aging assessment result to obtain an optimal inspection path, and inspecting the photovoltaic power station according to the optimal inspection path.

[0025] In step S11 , power station location information, spectral reflectance data and initial geographic fence are acquired.

[0026] In one implementation, the power station location information is obtained through differential GPS technology, and a dual-frequency receiver is used for spatial positioning with centimeter-level accuracy. The data is stored in the WGS-84 coordinate system as a three-dimensional coordinate group containing longitude, latitude and elevation. The format adopts the GeoJSON standard structure and comes with metadata fields such as positioning accuracy (HDOP≤1.5) and acquisition timestamp. Each set of data corresponds to the physical location of a specific inverter or combiner box.

[0027] In one embodiment, spectral reflectance data is collected by a hyperspectral imager in the visible light-near infrared band (400-2500nm), carried on a patrol drone or remote sensing satellite platform, and generated into 256-band reflectance cube data after radiation calibration. The stored content includes the original DN value, the surface reflectance after atmospheric correction, the solar altitude angle at the time of shooting, and the sensor attitude parameters. It is encapsulated in a hierarchical data format (HDF5) and associated with the GPS position index. The spatial resolution is set to 0.5-2 meters based on the payload performance.

[0028] In one implementation, the initial geographic fence is based on sub-meter satellite remote sensing images, and the photovoltaic array contour is extracted using an improved Canny edge detection algorithm. The OpenStreetMap vector road network data is combined for spatial alignment to generate an ESRI Shapefile file containing a polygon vertex coordinate sequence. The boundary accuracy is controlled within the range of ±5 meters, and 10-15 meters of redundant space is allowed to be compatible with subsequent optimization algorithms. At the same time, the fence confidence level (level 1-3) and the topological relationship attributes of adjacent areas are marked.

[0029] In step S12, aging analysis is performed based on the spectral reflectance data to obtain material aging characteristics.

[0030] In one implementation, a linear regression analysis is performed based on the spectral reflectance data and a preset first wavelength range to obtain a spectral slope; Performing Gaussian fitting based on the spectral reflection data and a preset second wavelength range to obtain a wavelength reflection function; Performing a second-order derivative calculation based on the wavelength reflection function to obtain a reflection second-order derivative; The wavelength corresponding to the maximum value of the second-order derivative of the reflection is taken as the reflection peak position; constructing a spectral matrix according to the spectral reflectance data; Performing principal component analysis on the spectral matrix to obtain a dimension-reduced spectral matrix; The pollution index is calculated according to the following formula: in, represents the pollution index, Indicates the dimension number of the reduced spectral matrix, represents the total number of dimensions of the reduced spectral matrix, express The weight coefficient of the dimension, Represents the reduced spectral matrix The value of the dimension; The material aging characteristics include the spectral slope, the reflection peak position and the pollution index.

[0031] It is worth noting that the first wavelength range is the visible light band (i.e., 450-700nm). In the linear regression analysis, the collected original spectral data must first be preprocessed: the Savitzky-Golay filtering algorithm is used to eliminate random noise interference, and the cubic spline interpolation method is used to resample the non-uniformly sampled spectral data to a standardized wavelength sequence with an interval of 1nm. Subsequently, 20 characteristic wavelength points in this band (such as 450nm, 475nm, 500nm and other key nodes) are selected to construct a wavelength-reflectivity scatter plot, and the weighted least squares method is used for linear regression modeling-the weight coefficient is dynamically adjusted according to the signal-to-noise ratio of each wavelength point to ensure that the high signal-to-noise ratio data points contribute more to the fitting results. The slope parameter obtained by fitting represents the average change in reflectivity for each increase of 1nm wavelength: a negative slope indicates that the reflectivity is attenuated in the long-wave direction, and a positive slope reflects the trend of enhanced reflection.

[0032] It is worth noting that the second wavelength range is the near-infrared band (this method is set to 900-1700nm). The Gaussian fitting process is as follows: First, the original data is baseline corrected, and the local weighted scatter smoothing (LOESS) algorithm is used to eliminate the baseline drift caused by the thermal noise of the detector. The abnormal reflection values ​​caused by surface contamination are eliminated by the adaptive threshold method, and the proportion of valid data points must be retained to be ≥95%. The preprocessed spectral data is normalized to unify the reflectance value range to the range of 0-1. The wavelength reflectance function is fitted using a seven-parameter Gaussian model: in, The wavelength is The wavelength reflectance function is represents the exponential function, represents the wavelength, Represents the coefficient of function parameter; The second-order derivative calculation uses the five-point method to perform discrete second-order derivative calculations: in, represents the second derivative of the wavelength reflectance function, represents the wavelength, Indicates the instrument resolution (value is 5nm), The wavelength is The wavelength reflectance function is The wavelength is The wavelength reflectance function is The wavelength is The wavelength reflectance function is The wavelength is The wavelength reflectance function is The wavelength is The wavelength reflectance function.

[0033] In one embodiment, the spectral reflectance data of multiple samples are first arranged in rows to form an original data matrix X, which has m rows (number of samples) and 251 columns (corresponding to a wavelength range of 450-700nm, with an interval of 1nm). Each element represents the reflectance value of the corresponding sample at a specific wavelength point. Data standardization is then performed: the mean and standard deviation of the reflectance values ​​at each wavelength point are calculated respectively, and then each value is subtracted from the mean of the column and divided by the standard deviation to ensure that each wavelength dimension has zero mean and unit variance. Next, the covariance matrix of the standardized data is calculated by multiplying the transposed standardized matrix with itself and dividing it by m-1. The covariance matrix is ​​eigen-decomposed to obtain the eigenvalues ​​and their corresponding eigenvectors (principal component directions) arranged in descending order of size. The number of principal components is determined by the cumulative variance contribution rate. n , select the minimum number of principal components that makes the cumulative contribution rate reach more than 95%. Finally, project the standardized data onto the selected front n The spectral matrix with reduced dimensions is obtained by matrix multiplication. The number of rows remains the same as the number of samples, and the number of columns is reduced to n principal component dimensions.

[0034] It is worth noting that the principal component dimensions of the reduced spectral matrix are 3, corresponding to the pollution characteristics of VOCs, heavy metals, and particulate matter. The larger the value of the pollution index, the higher the degree of pollution of the photovoltaic power station.

[0035] In step S13, the material aging characteristics are input into a pre-trained aging assessment model, and an aging assessment result is output.

[0036] In one implementation, an aging assessment model is constructed based on historical aging characteristics and historical aging results, and the model is trained. When the number of training times reaches a set upper limit or the loss function of the detection model meets the conditions, the training is determined to be completed, and a trained model is obtained; the material aging characteristics are input into the trained model to obtain an aging assessment result.

[0037] In one embodiment, the aging assessment model is based on an improved convolutional neural network. The input features include: spectral slope, which characterizes the rate of change of the chemical structure of the material surface (the mean of the first-order derivative of the reflectance spectrum in the wavelength range of 400-700nm); (2) reflectance peak position, which refers to the central wavelength offset of the characteristic reflectance peak of the material (accuracy of ±0.5nm); (3) contamination index, which calculates the proportion of foreign atom concentration through EDX elemental analysis. A deep residual network (ResNet-18 variant) is used for feature fusion and aging prediction. The network architecture contains 4 residual blocks and an SE channel attention module is embedded in the third layer to enhance the weight learning of key aging features. During training, a sample set containing 50,000 sets of historical aging data is used. 200 rounds of iterative training are performed through the AdamW optimizer (learning rate 3e-4) with a cross entropy loss function with class weights (weight ratio 1:2:3:4:5). The training is terminated when the accuracy of the validation set fluctuates less than 0.5% for 15 consecutive rounds. The final output includes: ① aging level (1-5, 5 is severe aging), which is divided according to the weighted distance of the feature deviation from the baseline value; ② estimated service life (unit: month), which is output through full connection layer regression and takes the lower limit of the 95% confidence interval. For example, the output "Level 3 (confidence 82%), remaining service life 23±5 months" means that the material has entered the accelerated aging stage and needs to be tested more strictly.

[0038] In step S14, spatial clustering is performed according to the power station location information to obtain power station clustering data.

[0039] In one implementation, the power station location information is converted into a coordinate system to obtain plane location data; a neighborhood radius and a minimum number of core points are obtained; the plane location data are traversed; points whose number of coordinate points within the neighborhood radius is greater than the minimum number of core points are marked as core points; coordinate points within the neighborhood of the core point that do not meet the core standard are marked as boundary points, and the boundary points are incorporated into the core point cluster of the core point; coordinate points that are not incorporated into any core point cluster are marked as noise points; wherein the power station clustering data includes the distribution of the core points, the boundary points, and the noise points.

[0040] It is worth noting that the core standard is that the points whose coordinate points within the neighborhood radius are greater than the minimum number of core points are core points.

[0041] It is worth noting that spatial clustering is implemented using an improved DBSCAN algorithm. When it is implemented specifically: first, the geographical location (longitude and latitude) of the power station is converted into plane rectangular coordinate system data through UTM projection to eliminate the influence of the earth's curvature; the neighborhood radius is set to 5 kilometers and the minimum number of core points is set to 3. When clustering is performed, the algorithm traverses all coordinate points, and when there are at least 3 adjacent power stations in the neighborhood of a point, it is marked as a core point, and then all adjacent points within 5 kilometers of the core point (including boundary points that do not meet the core standard) are classified into the same cluster; this process is recursively extended until no new members can be added, and finally three types of distribution are formed: core point clusters (representing high-density power station groups), boundary points (located at the edge of the cluster) and noise points (isolated power stations, more than 5 kilometers away from the nearest cluster).

[0042] In step S15, iterative optimization is performed based on the power station clustering data and the initial geo-fence to obtain an optimized geo-fence.

[0043] In one embodiment, an initial convex hull fence is generated based on the power station clustering data; a spatial topological analysis is performed based on the initial geo-fence and the power station clustering data to obtain an abnormal fence; the abnormal fence is optimized according to a preset fence optimization rule to obtain an updated geo-fence; the updated geo-fence and the initial convex hull fence are merged to obtain an optimized geo-fence.

[0044] In one embodiment, the spatial topological analysis is performed based on the initial geographic fence and the power station clustering data to obtain an abnormal fence, including: when there is a core point outside the initial geographic fence, it is determined to be an out-of-bounds component fence; when there is no core point inside the initial geographic fence, it is determined to be a void area fence; the power station density is calculated based on the initial geographic fence and the power station clustering data to obtain the power station density; when the peak difference of the power station density within the initial geographic fence is greater than a preset density threshold, it is determined to be a density mutation area fence.

[0045] It is worth noting that in the iterative optimization process of geo-fences, the initial convex hull fence is first constructed using the Graham scanning algorithm, which generates the minimum convex hull boundary based on the core point clusters generated by DBSCAN clustering. Then, three types of abnormal fences are identified through three-dimensional spatial topological relationship analysis for targeted processing: It is worth noting that the optimization of the fence for out-of-bounds components occurs when there is a high-density core point cluster outside the initial fence. For example, when some points of a discretely distributed power station group exceed the boundary, the discrete points will be included in the new boundary through polygon vertex expansion or bounding box extension operations. For the fence of the empty area formed by the original fence mistakenly including the area where the station cannot be built (such as the lake protection area), the redundant space range is cut off by spatial Boolean operation. In the special case of the fence in the density mutation area, when there is a significant difference in the power station density of 15 / square kilometer on the east side and 3 / square kilometer on the west side in a certain area, a transition sub-fence will be constructed based on the Voronoi diagram segmentation technology, and a density gradient threshold of ±20% will be set to trigger the dynamic reconstruction mechanism.

[0046] It is worth noting that in the final stage, the geometric joint operation function of the GEOS library is used to merge the optimized abnormal fence with the initial convex hull fence to form a seamless closed optimized geographic fence. This fence not only completely covers all core point clusters, but also accurately reflects the spatial distribution characteristics of the power station and effectively eliminates geographical redundant areas.

[0047] In step S16, inspection path planning is performed according to the optimized geographic fence and the aging assessment result to obtain an optimal inspection path, and the photovoltaic power station is inspected according to the optimal inspection path.

[0048] In one embodiment, inspection areas are divided according to the optimized geo-fence; high-risk components are marked according to the aging assessment results and the optimized geo-fence; a basic inspection path that can traverse all the inspection areas is initialized; key inspection points are generated according to the high-risk components; the key inspection points are inserted into the basic inspection path to obtain an updated inspection path; and the updated inspection path is smoothed to obtain an optimal inspection path.

[0049] In one implementation, the photovoltaic power station is divided into logical inspection units based on the optimized geographical fence polygon boundary using Voronoi diagram segmentation technology. Each unit is mapped to the geographical coordinates of the equipment through spatial indexing, and the color temperature heat map generated by the aging assessment result is superimposed to automatically identify high-risk components with corrosion rates exceeding the threshold (such as annual corrosion volume > 0.5mm) or significant electrical performance degradation (such as conversion efficiency reduction > 15%), forming a three-dimensional labeled point cloud data set containing geographical coordinates, risk levels, and detection parameters.

[0050] In one implementation, an improved genetic algorithm is used to construct the initial inspection path. The algorithm sets dual constraints: the spatial constraint requires that the path completely covers the minimum outer rectangle of all inspection units, and the time constraint limits the duration of a single operation to no more than 8 hours. A weighted graph model is generated through equipment positioning data. The node weights include terrain undulation (DEM elevation data) and obstacle distribution density. The edge weights are calculated based on the UAV flight energy consumption model, and the final output is a Hamiltonian circuit that meets the full coverage requirements.

[0051] In one implementation, a two-layer detection strategy is initiated for marked high-risk component clusters: in-situ fixed-point detection is performed for primary risk points (such as insulation resistance <2MΩ) to generate a spherical detection area with a diameter of 50 meters; for secondary risk points (such as backplane cracking rate >30%), a scanning path along the component array is set. The shortest polyline distance between the basic path and the risk point is calculated using the Dijkstra algorithm, and the detection area is seamlessly embedded in the original path using the cubic spline interpolation method to ensure the continuity of the curvature of the newly added detection point and the basic path.

[0052] In one implementation, the Pareto frontier analysis method is introduced to handle multi-dimensional optimization objectives: in the energy consumption dimension, the lithium battery consumption rate of different flight segments is calculated according to the UAV power model; in the time dimension, the weight of the influence of changes in lighting conditions on visible light inspection is considered; in the detection accuracy dimension, the hovering time is adjusted in combination with the wind speed disturbance model. The NSGA-II algorithm is used for non-dominated sorting, and finally the path plan with the best comprehensive score is selected. The plan has the following characteristics: the proportion of high-risk area detection time is ≥40%, the total flight range fluctuation rate is <15%, and the length of redundant segments is controlled within 5% of the total path.

[0053] In one implementation, a path smoothing engine based on NURBS curve theory is deployed, which achieves three levels of smoothing through control point optimization: the global level eliminates sharp turns in the path and increases the turning radius to a safe threshold (≥10 meters); the local level adopts a spiral progressive approach strategy in equipment-intensive areas; the micro level dynamically adjusts the waypoint spacing based on real-time meteorological data. Finally, a three-dimensional track file that meets the RTK positioning accuracy is output, and a digital twin scene is built through the Unity3D engine. Monte Carlo simulation is used to verify the feasibility of the path, ensuring that more than 98% of the detection point cloud errors are controlled within the range of ±0.3 meters.

[0054] In one implementation, an evaluation index system is established that includes the equipment body status (such as the backplane crack area, the proportion of dark spots in the EL image), environmental corrosion parameters (salt spray deposition rate, acid rain pH value) and electrical performance degradation (IV curve distortion, insulation impedance drop rate). The analytic hierarchy process is used to determine the weight of each indicator, for example, backplane defects account for 35%, electrical performance accounts for 40%, and environmental factors account for 25%. The equipment is divided into four levels of risk through fuzzy comprehensive evaluation: Level 1 (comprehensive score ≥ 85 points) requires emergency treatment, Level 2 (70-84 points) is key monitoring, Level 3 (50-69 points) is routine testing, and Level 4 (<50 points) is sampling inspection. Taking a coastal power station as an example, 12.7% of its components reached the second-level risk due to salt spray corrosion, and the inspection frequency needs to be increased to twice a week.

[0055] In one implementation, a "three-in-one" inspection is implemented for high-risk equipment: deploy infrared thermal imagers for all-weather temperature field monitoring (sampling interval ≤ 10 minutes); configure drones equipped with EL detection modules to perform fixed-point hovering imaging (hovering accuracy ±0.1m); and arrange manual handheld IV testers for contact diagnosis. Medium-risk equipment uses mobile IV curve scanning vehicles for daytime inspections (coverage rate 100%). Low-risk equipment uses intelligent cleaning robots equipped with simple sensors to achieve monthly census, and the inspection time is reduced from the conventional 45 minutes / group to 8 minutes / group.

[0056] In summary, the present invention discloses a multimodal collaborative photovoltaic power station inspection method, which aims to improve the inspection efficiency of photovoltaic power stations by integrating multiple data sources and applying data analysis methods. The method first obtains the power station location information, spectral reflection data and initial geographic fence to provide a comprehensive data basis for subsequent analysis. In the aging analysis stage, the spectral reflection data is used to accurately identify the aging characteristics of the material, such as spectral slope, reflection peak position and pollution index. Compared with the traditional time-based or experience-based evaluation, this method is more scientific and accurate, and can effectively locate the area that needs attention. Then, these material aging characteristics are input into the pre-trained aging evaluation model to output the aging evaluation results. This process not only improves the accuracy of the evaluation, but also accelerates the evaluation process. Since the model is trained based on a large amount of historical data, it can quickly provide reliable prediction results on new data. In addition, spatial clustering is performed according to the power station location information to obtain power station clustering data, and iterative optimization is performed in combination with the initial geographic fence to obtain the optimized geographic fence, which helps to improve the accuracy and practicality of the geographic fence, so that the inspection activities can cover the target area more efficiently, while reducing unnecessary repeated inspections. In terms of inspection route planning, the optimal inspection route is generated based on the optimized geographic fence and aging assessment results, which fully considers the actual status and distribution of the power station to ensure the targetedness and effectiveness of the inspection work.

[0057] For example, when processing spectral reflectance data, a linear regression analysis is first performed based on the preset first wavelength range (450-700nm) to obtain the spectral slope; then a Gaussian fit is performed based on the second wavelength range (900-1700nm) to obtain the wavelength reflectance function, and the second-order derivative is calculated to determine the reflection peak position. These steps work together to effectively capture signs of aging on the surface of the material. Afterwards, by performing principal component analysis (PCA) on the spectral matrix, the data dimension can be reduced, simplifying subsequent analysis while retaining key information. It is worth mentioning that throughout the process, the concept of a contamination index is also introduced to quantify the degree of contamination of photovoltaic modules, which is crucial for judging the working status of the modules.

[0058] As one of the cores of the entire system, the construction and training process of the aging assessment model also reflects a high degree of technical content. The model is built based on historical aging characteristics and historical aging results, and is trained repeatedly until specific conditions are met. After training, the model can receive new material aging feature inputs and quickly give aging assessment results. This machine learning-based method greatly improves the speed and accuracy of traditional manual assessments. At the same time, the application of spatial clustering algorithms has also brought significant improvements to inspection work. Through the DBSCAN algorithm or its variants, power station units with similar geographical locations and similar states can be classified together to form an easy-to-manage clustered data set. Then, combined with the information of the initial geo-fence, the design of the geo-fence is further optimized to ensure that the inspection plan covers all necessary areas without over-expansion.

[0059] Finally, in the inspection route planning stage, the system will divide the inspection area according to the optimized geographic fence, and mark the high-risk components in combination with the aging assessment results. The initialized basic inspection path will be adjusted on this basis, adding key inspection points for high-risk components, and finally generating the optimal inspection path after smoothing. This process not only takes into account physical distance and terrain factors, but also incorporates complex variables such as drone flight energy consumption models, striving to achieve the most efficient inspection route design. Overall, the present invention has shown significant advantages in improving the inspection efficiency of photovoltaic power stations. It can not only more accurately identify components that need maintenance, but also optimize inspection routes to ensure that each key point can be properly inspected.

[0060] Reference Figure 2 The second embodiment of the present invention provides a multi-modal collaborative photovoltaic power station inspection system, including: A data acquisition module, used to obtain power station location information, spectral reflectance data and initial geo-fence; An aging analysis module, used to perform aging analysis based on the spectral reflectance data to obtain material aging characteristics; An aging assessment module, used for inputting the material aging characteristics into a pre-trained aging assessment model and outputting an aging assessment result; A power station clustering module, used for performing spatial clustering according to the power station location information to obtain power station clustering data; A fence optimization module, used for iteratively optimizing according to the power station clustering data and the initial geo-fence to obtain an optimized geo-fence; A path planning module is used to plan an inspection path according to the optimized geographic fence and the aging assessment result, obtain an optimal inspection path, and inspect the photovoltaic power station according to the optimal inspection path.

[0061] It should be noted that a multimodal collaborative photovoltaic power station inspection system provided in an embodiment of the present invention is used to execute all process steps of a multimodal collaborative photovoltaic power station inspection method in the above embodiment. The working principles and beneficial effects of the two correspond one to one, and thus will not be elaborated upon.

[0062] The embodiment of the present invention further provides an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a data acquisition program. When the processor executes the computer program, the steps in the above-mentioned embodiments of the multi-modal collaborative photovoltaic power station inspection method are implemented, such as Figure 1 Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are realized, such as the data acquisition module.

[0063] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the electronic device.

[0064] The electronic device may be a computing device such as a desktop computer, a notebook, a PDA, and a smart tablet. The electronic device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the above components are merely examples of electronic devices and do not constitute a limitation on the electronic device. The electronic device may include more or fewer components than the above components, or may combine certain components, or different components. For example, the electronic device may also include input and output devices, network access devices, buses, etc.

[0065] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the electronic device, and uses various interfaces and lines to connect various parts of the entire electronic device.

[0066] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the electronic device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0067] Wherein, if the module / unit integrated in the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0068] It should be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without paying any creative effort.

[0069] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-modal collaborative photovoltaic power station inspection method, characterized in that: include: Obtain power station location information, spectral reflectance data, and initial geo-fence; Performing aging analysis based on the spectral reflectance data to obtain material aging characteristics; Inputting the material aging characteristics into a pre-trained aging assessment model and outputting an aging assessment result; Performing spatial clustering according to the power station location information to obtain power station clustering data; Iteratively optimizing the power station clustering data and the initial geo-fence to obtain an optimized geo-fence; Inspection paths are planned according to the optimized geographic fence and the aging assessment result to obtain an optimal inspection path, and the photovoltaic power station is inspected according to the optimal inspection path.

2. The multi-modal collaborative photovoltaic power station inspection method according to claim 1, characterized in that: The performing aging analysis according to the spectral reflectance data to obtain material aging characteristics includes: Performing a linear regression analysis based on the spectral reflectance data and a preset first wavelength range to obtain a spectral slope; Performing Gaussian fitting based on the spectral reflection data and a preset second wavelength range to obtain a wavelength reflection function; Performing a second-order derivative calculation based on the wavelength reflection function to obtain a reflection second-order derivative; The wavelength corresponding to the maximum value of the second-order derivative of the reflection is taken as the reflection peak position; constructing a spectral matrix according to the spectral reflectance data; Performing principal component analysis on the spectral matrix to obtain a dimension-reduced spectral matrix; The pollution index is calculated according to the following formula: in, represents the pollution index, Indicates the dimension number of the reduced spectral matrix, represents the total number of dimensions of the reduced spectral matrix, express The weight coefficient of the dimension, Represents the reduced spectral matrix The value of the dimension; The material aging characteristics include the spectral slope, the reflection peak position and the pollution index.

3. The multi-modal collaborative photovoltaic power station inspection method according to claim 1, characterized in that: The training process of the aging assessment model includes: An aging assessment model is constructed based on historical aging characteristics and historical aging results, and the model is trained. When the number of training times reaches a set upper limit or the loss function of the detection model meets the conditions, the training is considered complete, and the trained model is obtained; The material aging characteristics are input into the trained model to obtain an aging assessment result.

4. The multi-modal collaborative photovoltaic power station inspection method according to claim 1, characterized in that: The performing spatial clustering according to the power station location information to obtain power station clustering data includes: Performing coordinate system conversion on the power station location information to obtain plane location data; Get the neighborhood radius and minimum number of core points; traversing the plane position data; Marking points within the neighborhood radius whose number of coordinate points is greater than the minimum number of core points as core points; Marking coordinate points in the neighborhood of the core point that do not meet the core standard as boundary points, and merging the boundary points into the core point cluster of the core point; The coordinate points that are not incorporated into any core point cluster are marked as noise points; The power station clustering data includes the distribution of the core points, the boundary points and the noise points.

5. The multi-modal collaborative photovoltaic power station inspection method according to claim 1, characterized in that: The iterative optimization according to the power station clustering data and the initial geo-fence to obtain an optimized geo-fence includes: generating an initial convex hull fence according to the power station clustering data; Performing spatial topological analysis based on the initial geographic fence and the power station clustering data to obtain an abnormal fence; Optimizing the abnormal fence according to a preset fence optimization rule to obtain an updated geographic fence; The updated geo-fence and the initial convex hull fence are merged to obtain an optimized geo-fence.

6. The multi-modal collaborative photovoltaic power station inspection method according to claim 5, characterized in that: The performing of spatial topological analysis based on the initial geographic fence and the power station clustering data to obtain an abnormal fence includes: When there is a core point outside the initial geo-fence, it is determined to be an out-of-bounds component fence; When there is no core point in the initial geographic fence, it is determined to be an empty area fence; Calculating power station density according to the initial geographic fence and the power station clustering data to obtain power station density; When the peak value difference of the power station density within the initial geographic fence is greater than a preset density threshold, it is determined to be a density mutation area fence.

7. The multi-modal collaborative photovoltaic power station inspection method according to claim 1, characterized in that: The performing inspection path planning according to the optimized geographic fence and the aging assessment result to obtain an optimal inspection path, and inspecting the photovoltaic power station according to the optimal inspection path, includes: Divide the inspection area according to the optimized geo-fence; marking high-risk components based on the aging assessment results and the optimized geo-fencing; Initialize a basic inspection path that can traverse all the inspection areas; generating key inspection points according to the high-risk components; Inserting the key detection points into the basic inspection path to obtain an updated inspection path; The updated inspection path is smoothed to obtain an optimal inspection path.

8. A multi-modal collaborative photovoltaic power station inspection system, characterized in that: include: A data acquisition module, used to obtain power station location information, spectral reflectance data and initial geo-fence; An aging analysis module, used to perform aging analysis based on the spectral reflectance data to obtain material aging characteristics; An aging assessment module, used for inputting the material aging characteristics into a pre-trained aging assessment model and outputting an aging assessment result; A power station clustering module, used for performing spatial clustering according to the power station location information to obtain power station clustering data; A fence optimization module, used for iteratively optimizing according to the power station clustering data and the initial geo-fence to obtain an optimized geo-fence; A path planning module is used to plan an inspection path according to the optimized geographic fence and the aging assessment result, obtain an optimal inspection path, and inspect the photovoltaic power station according to the optimal inspection path.

9. An electronic device, characterized in that: The invention comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the multimodal collaborative photovoltaic power station inspection method as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the multi-modal collaborative photovoltaic power station inspection method according to any one of claims 1 to 7.

Images