Photovoltaic operation and maintenance management system based on multi-modal data

Abstract

The invention belongs to the technical field of photovoltaic operation and maintenance management, and particularly provides a photovoltaic operation and maintenance management system based on multi-modal data, and the system mainly comprises a basic data module which obtains the three-dimensional coordinates of a photovoltaic array and the installation inclination angle data of a photovoltaic module; the thermal field acquisition module is used for instructing the unmanned aerial vehicle to fly based on the three-dimensional coordinates of the photovoltaic array, and synchronously acquiring an infrared thermal map and environmental parameters to generate a string thermal field with the environmental parameters; the anomaly positioning module is used for calculating an inclination angle difference of adjacent assemblies based on installation inclination angle data, generating an inclination angle gradient distribution diagram and positioning coordinates of a local temperature rise anomaly region by combining the group string thermal field; and the temperature rise correction module is used for performing heat loss compensation correction on the coordinates of the local temperature rise abnormal region and analyzing the real temperature rise amplitude. According to the invention, real hot spots are accurately identified, the diagnosis false alarm rate is reduced, the operation and maintenance resource configuration is optimized, the fault response speed and precision are improved, and reliable operation and operation and maintenance efficiency of the photovoltaic system are guaranteed.

Description

Technical Field

[0001] This application belongs to the field of photovoltaic operation and maintenance management technology, and specifically relates to a photovoltaic operation and maintenance management system based on multimodal data. Background Technology

[0002] In photovoltaic operation and maintenance management, the industry generally uses drones to collect infrared thermal images, integrates environmental parameters to construct the thermal field of photovoltaic strings, and correlates them with real-time power generation data to carry out fault diagnosis in order to support operation and maintenance decisions.

[0003] However, in distributed rooftop photovoltaic (PV) scenarios, PV arrays are often arranged across different roof slopes, resulting in inherent deviations in the tilt angle of the installed modules. The above-mentioned solutions do not fully explore the intrinsic relationship between the tilt angle deviation of the installed modules and the heat dissipation characteristics of the strings. When the tilt angle difference between adjacent modules reaches a certain threshold and is in a weak heat dissipation environment, the uneven heat dissipation caused by the tilt angle deviation will distort the thermal field analysis logic, causing a mismatch between the hot spot judgment threshold and the actual operating conditions, resulting in an increased false alarm rate for hot spot diagnosis. This, in turn, leads to misallocation of operation and maintenance resources, delayed fault response, and restricts the reliable operation and maintenance efficiency of the PV system. Summary of the Invention

[0004] This application provides a photovoltaic operation and maintenance management system based on multimodal data, which effectively solves the problems of hot spot false alarms, operation and maintenance mismatch, and response lag caused by the deviation of the component tilt angle due to the roof slope in the existing distributed roof photovoltaic system and the failure to consider the correlation between tilt angle and heat dissipation in weak heat dissipation environment. It realizes accurate identification of real hot spots, reduces the false alarm rate of diagnosis, optimizes the allocation of operation and maintenance resources, improves the speed and accuracy of fault response, and ensures the reliable operation and operation and maintenance efficiency of photovoltaic system.

[0005] To achieve the above objectives, this application adopts the following technical solution:

[0006] Firstly, this application provides a photovoltaic operation and maintenance management system based on multimodal data, including:

[0007] Basic data module: Acquires the three-dimensional coordinates of the photovoltaic array and the installation tilt angle data of the photovoltaic modules.

[0008] Thermal field acquisition module: Based on the three-dimensional coordinate command of the photovoltaic array, the UAV flies and simultaneously acquires infrared thermal images and environmental parameters to generate string thermal fields with environmental parameters.

[0009] Anomaly location module: Calculates the tilt angle difference between adjacent components based on installation tilt angle data, generates a tilt angle gradient distribution map, and locates the coordinates of local temperature rise anomaly areas by combining the string thermal field.

[0010] Temperature rise correction module: Performs heat loss compensation correction on the coordinates of the local temperature rise anomaly area and analyzes the true temperature rise amplitude.

[0011] Fault verification module: acquires real-time power generation data, correlates the actual temperature rise with the real-time power generation data, and verifies and outputs the fault component number.

[0012] Path planning module: Generates a three-dimensional inspection path based on the three-dimensional spatial distribution of the confirmed fault component numbers.

[0013] Path planning module: instructs the UAV to execute the three-dimensional inspection path and obtain maintenance feedback, and corrects the judgment rules for locating the coordinates of the local temperature rise anomaly area based on the maintenance feedback, and outputs a dynamic diagnostic model.

[0014] Furthermore, based on the three-dimensional coordinate commands of the photovoltaic array, the UAV flies while simultaneously acquiring infrared thermal images and environmental parameters to generate a string thermal field with environmental parameters, including:

[0015] Plan the UAV flight path based on the three-dimensional coordinates of the photovoltaic array.

[0016] The command instructs the drone to fly along the trajectory and obtains the drone's real-time spatial coordinates.

[0017] The infrared thermal imager is triggered at the real-time spatial coordinates to collect raw temperature distribution data.

[0018] Normalize the raw temperature distribution data and output a standardized heatmap.

[0019] It integrates standardized thermal maps with real-time wind speed and ambient temperature from weather stations to output string thermal fields with environmental parameters.

[0020] Furthermore, based on the installation tilt angle data, the tilt angle difference between adjacent components is calculated to generate a tilt angle gradient distribution map. Combined with the string thermal field, the coordinates of local temperature rise anomaly areas are located, including:

[0021] Extract the tilt angle values ​​of adjacent components from the installation tilt angle data.

[0022] Calculate the absolute value of the difference between adjacent dip angle values ​​to generate a dip angle gradient data matrix.

[0023] Convert the tilt gradient data matrix into a spatial distribution map and output the tilt gradient distribution map.

[0024] Identify the coordinates of blocks in the tilt gradient distribution map where the difference is continuously greater than a preset temperature threshold, and mark the heat dissipation blockage area.

[0025] Map the heat dissipation blockage area markers to the corresponding positions in the string thermal field, and output the coordinates of the local temperature rise anomaly area.

[0026] Furthermore, the coordinates of the local temperature rise anomaly area are corrected for heat loss compensation to analyze the true temperature rise amplitude, including:

[0027] The temperature values ​​of the corresponding components are extracted from the coordinates of the local temperature rise anomaly area, and a candidate hot spot temperature set is output; the theoretical heat loss is calculated based on the real-time wind speed and ambient temperature.

[0028] The theoretical heat loss is subtracted from the candidate hotspot temperature set, and the corrected temperature value is output.

[0029] Obtain the average temperature of healthy components in the same string, compare the corrected temperature value with the average temperature, and calculate the actual temperature rise.

[0030] Furthermore, by correlating the actual temperature rise with the real-time power generation data, the faulty component number is verified and output, including:

[0031] Components whose actual temperature rise exceeds the preset temperature rise threshold are selected, and a list of components to be verified is generated.

[0032] For the list of components to be verified, retrieve the inverter power generation during the same period and output the target power time-series curve.

[0033] Calculate the decay rate of the target power time-series curve relative to the healthy component, and generate an electrothermal correlation factor.

[0034] When the electrothermal correlation factor remains below the preset fault threshold, the fault component number is output to confirm the fault.

[0035] Furthermore, a three-dimensional inspection path is generated based on the three-dimensional spatial distribution of the confirmed fault component numbers, including:

[0036] Associate and confirm the three-dimensional geographic coordinates of the faulty component number to construct a spatial topology map of the fault point.

[0037] Based on the maximum flight range of the drone, the spatial topology map is segmented to divide the emergency maintenance sub-region.

[0038] Obstacle avoidance paths are generated by overlaying roof safety passage parameters on the emergency maintenance sub-area, resulting in a three-dimensional inspection path.

[0039] Furthermore, based on the criterion rule for correcting the coordinates of local temperature rise anomaly areas through maintenance feedback, a dynamic diagnostic model is output, including:

[0040] Perform maintenance according to the three-dimensional inspection path and obtain the thermal field data after maintenance.

[0041] Compare the post-maintenance thermal field data with the actual temperature rise to calculate the temperature rise elimination rate.

[0042] The tilt difference determination rule used to generate the tilt gradient distribution map is invoked, the tilt change sensitivity parameter is adjusted according to the temperature rise elimination rate, and a dynamic diagnostic model is output.

[0043] Furthermore, the dip gradient data matrix is ​​converted into a spatial distribution map, and the dip gradient distribution map is output, including:

[0044] Spatial interpolation is performed on the tilt gradient data matrix to fill in the gaps between components and generate a continuous gradient field.

[0045] Pseudo-color mapping based on a continuous gradient field is used to identify the temperature difference intensity, resulting in a tilt gradient distribution map.

[0046] Furthermore, the theoretical heat loss is calculated based on real-time wind speed and ambient temperature, including:

[0047] The surface convective heat transfer coefficient is calculated based on real-time wind speed and ambient temperature, and the heat transfer coefficient value is output.

[0048] The heat loss power is calculated based on the heat transfer coefficient value and the standard area of ​​the component, and the heat loss power value is output.

[0049] The heat loss power value is converted into a temperature compensation equivalent, and the theoretical heat loss is output.

[0050] Furthermore, by comparing the post-maintenance thermal field data with the actual temperature rise, the temperature rise elimination rate is calculated, including:

[0051] The steady-state temperature of the target component is extracted from the thermal field data after maintenance, and the remeasured temperature value is output.

[0052] By correlating the remeasured temperature values ​​with the original actual temperature rise at the same location, a temperature rise change comparison group is generated.

[0053] Calculate the relative attenuation ratio of the temperature rise change comparison group and output the temperature rise elimination rate.

[0054] The map construction and decision support module generates a three-dimensional map of volatiles across the entire domain based on the volatiles dataset, and outputs decision support data in conjunction with the pharmacological knowledge base.

[0055] Secondly, this application provides a photovoltaic operation and maintenance management method based on multimodal data, including: acquiring the three-dimensional coordinates of the photovoltaic array and the installation tilt angle data of the photovoltaic modules.

[0056] The drone flies based on the three-dimensional coordinate commands of the photovoltaic array, and simultaneously collects infrared thermal images and environmental parameters to generate string thermal fields with environmental parameters.

[0057] Based on the installation tilt angle data, the tilt angle difference between adjacent components is calculated, a tilt angle gradient distribution map is generated, and the coordinates of the local temperature rise anomaly area are located by combining the string thermal field.

[0058] The coordinates of the local temperature rise anomaly area are corrected for heat loss compensation to analyze the true temperature rise amplitude.

[0059] Obtain real-time power generation data, correlate the actual temperature rise with the real-time power generation data, and verify and output the fault component number.

[0060] A three-dimensional inspection path is generated based on the three-dimensional spatial distribution of the confirmed fault component numbers.

[0061] The command drone executes the three-dimensional inspection path and obtains maintenance feedback. Based on the maintenance feedback, the criteria rule for locating the coordinates of the local temperature rise anomaly area is corrected in reverse, and a dynamic diagnostic model is output.

[0062] Thirdly, this application provides a photovoltaic operation and maintenance management device based on multimodal data, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program to implement the steps of the photovoltaic operation and maintenance management method based on multimodal data.

[0063] Fourthly, this application provides a storage medium storing computer program instructions, which are read and executed by a processor to perform the steps of a photovoltaic operation and maintenance management method based on multimodal data.

[0064] Fifthly, this application provides a computer program product, including a computer program or instructions, wherein when the computer program or instructions are executed by a processor, the steps of a photovoltaic operation and maintenance management method based on multimodal data are implemented.

[0065] The beneficial effects of this application are:

[0066] This application employs a technical solution that integrates the three-dimensional coordinates of the photovoltaic array with the tilt angle data of the components. Through thermal field construction, tilt angle gradient analysis, temperature rise correction, and power correlation verification to generate a three-dimensional inspection path, it effectively solves the problems of hot spot false alarms, operation and maintenance mismatch, and response lag caused by the tilt angle deviation of components due to the roof slope in existing distributed roof photovoltaic systems and the lack of consideration of the correlation between tilt angle and heat dissipation in weak heat dissipation environments. It achieves accurate identification of real hot spots, reduces the false alarm rate of diagnosis, optimizes the allocation of operation and maintenance resources, improves the speed and accuracy of fault response, and ensures the reliable operation and operation and maintenance efficiency of the photovoltaic system.

[0067] Other features and advantages of this application will be set forth in the following description and will be apparent in part from the description or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures pointed out in the description and the accompanying drawings. Attached Figure Description

[0068] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0069] Figure 1 A schematic diagram of the photovoltaic operation and maintenance management system based on multimodal data of this application is shown. Detailed Implementation

[0070] To address the problems raised in the background technology, this application employs a multimodal data approach that integrates the three-dimensional coordinates of the photovoltaic array, the tilt angle of the modules, infrared thermal images, environmental parameters, and real-time power generation. This approach sequentially constructs a string thermal field with environmental parameters, calculates the tilt angle difference between adjacent modules to generate a gradient distribution map, locates and corrects local abnormal temperature rise areas to analyze the true temperature rise, correlates electrical performance to verify faulty modules, and finally generates a three-dimensional inspection path.

[0071] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0072] In some embodiments, such as Figure 1 As shown, this application provides a photovoltaic operation and maintenance management system based on multimodal data, including:

[0073] Basic data module: Acquires the three-dimensional coordinates of the photovoltaic array and the installation tilt angle data of the photovoltaic modules.

[0074] Thermal field acquisition module: Based on the three-dimensional coordinate command of the photovoltaic array, the UAV flies and simultaneously acquires infrared thermal images and environmental parameters to generate string thermal fields with environmental parameters.

[0075] Anomaly location module: Calculates the tilt angle difference between adjacent components based on installation tilt angle data, generates a tilt angle gradient distribution map, and locates the coordinates of local temperature rise anomaly areas by combining the string thermal field.

[0076] Temperature rise correction module: Performs heat loss compensation correction on the coordinates of the local temperature rise anomaly area and analyzes the true temperature rise amplitude.

[0077] Fault verification module: acquires real-time power generation data, correlates the actual temperature rise with the real-time power generation data, and verifies and outputs the fault component number.

[0078] Path planning module: Generates a three-dimensional inspection path based on the three-dimensional spatial distribution of the confirmed fault component numbers.

[0079] Path planning module: instructs the UAV to execute the three-dimensional inspection path and obtain maintenance feedback, and corrects the judgment rules for locating the coordinates of the local temperature rise anomaly area based on the maintenance feedback, and outputs a dynamic diagnostic model.

[0080] Among them, obtaining the three-dimensional coordinates of the photovoltaic array represents the location data of the photovoltaic modules in real space obtained through surveying and mapping technology, including the longitude, latitude and altitude information of each module.

[0081] Photovoltaic module installation tilt angle data represents the real-time angle between the surface of the photovoltaic module and the horizontal plane, used to quantify the module's installation posture.

[0082] In some embodiments, the UAV flies based on the three-dimensional coordinates of the photovoltaic array, simultaneously acquiring infrared thermal images and environmental parameters to generate a string thermal field with environmental parameters, including:

[0083] S11. Plan the UAV flight path based on the three-dimensional coordinates of the photovoltaic array.

[0084] Specifically, by inputting the three-dimensional coordinates of the photovoltaic array, the A* pathfinding algorithm can be used to process the data to calculate the optimal flight path, and finally output the UAV flight track composed of a sequence of spatial coordinate points.

[0085] For example, in the process of calculating the optimal flight path, the following three points need to be ensured: first, to cover the airspace above the center points of all components; second, to achieve the shortest total path length, thereby reducing flight time; and third, to avoid obstacles between arrays.

[0086] S12. Instruct the UAV to fly along the flight path and obtain the UAV's real-time spatial coordinates.

[0087] Specifically, using the UAV flight path as the input path reference, the flight control system controls the UAV to fly along the path, and the UAV transmits spatial coordinate data in real time. The final output format is the UAV's real-time spatial coordinates in [timestamp, longitude, latitude, altitude].

[0088] S13. Trigger the infrared thermal imager to collect raw temperature distribution data at the real-time spatial coordinates.

[0089] Specifically, based on the real-time spatial coordinates of the UAV, when the UAV arrives at the preset track point, it triggers the onboard infrared thermal imager to capture a thermal image, and at the same time records the temperature value and corresponding spatial coordinates of each pixel in the thermal image; the final output matrix format is the original temperature distribution data of [x coordinate, y coordinate, temperature value].

[0090] S14. Perform normalization processing on the original temperature distribution data and output a standardized heatmap.

[0091] Calculate the dynamic temperature range: T range =T max -T min +ΔT; where T max T min These represent the highest and lowest temperatures in the original temperature distribution data, respectively, and ΔT represents the temperature buffer threshold, which can be set to 3-5℃.

[0092] By using linear normalization mapping, each temperature T i Convert to standard value T std .

[0093] Finally, the standard value T is... std The values ​​are mapped to grayscale levels, with 0 corresponding to low temperature and 255 corresponding to high temperature, generating a standardized heatmap.

[0094] S15. Integrates standardized thermal maps with real-time wind speed and ambient temperature from weather stations to output string thermal fields with environmental parameters.

[0095] Multi-source data spatiotemporal alignment technology can be used to fuse standardized heat maps, real-time wind speed from weather stations, and ambient temperature, outputting string thermal fields including x-coordinates, y-coordinates, photovoltaic module surface temperature, wind speed values, and ambient temperature values.

[0096] In some embodiments, the tilt angle difference between adjacent components is calculated based on the installation tilt angle data to generate a tilt angle gradient distribution map, which, combined with the string thermal field, locates the coordinates of local temperature rise anomaly areas, including:

[0097] S21. Extract the tilt angle values ​​of adjacent components from the installation tilt angle data.

[0098] Data can be sorted according to the physical location of photovoltaic modules, and the tilt angle values ​​of adjacent modules can be extracted.

[0099] S22. Calculate the absolute value of the difference between adjacent dip angle values ​​to generate a dip angle gradient data matrix.

[0100] After calculating the absolute value Δθ of the difference between adjacent tilt angles, a matrix is ​​generated based on the component positions.

[0101] Specifically, the photovoltaic array is arranged in a grid pattern, with each component having a fixed row and column position. A two-dimensional matrix that is completely isomorphic to the component layout is constructed, and the element value is the absolute value of the tilt angle difference between the component and the adjacent component on the right, thus obtaining the tilt angle gradient data matrix.

[0102] For example, suppose the array has M rows of components, with a maximum of N columns in each row. Create a matrix of M rows × (N-1) columns, where the position of the i-th row and j-th column of the matrix corresponds to the tilt angle difference between the j-th column and the j+1-th column of the i-th row of components.

[0103] When a component is at the end of a row, the corresponding matrix column is set to empty.

[0104] If the number of components in a row is less than N columns, leave the empty space blank.

[0105] S23. Convert the tilt gradient data matrix into a spatial distribution map and output the tilt gradient distribution map.

[0106] S24. Identify the coordinates of blocks in the tilt gradient distribution map where the difference is continuously greater than the preset temperature threshold, and mark the heat dissipation blockage area.

[0107] Specifically, identifying the coordinates of blocks in the tilt gradient distribution map where the difference is continuously greater than a preset temperature threshold can be done by identifying red blocks in the map whose tilt gradient satisfies θ≥5℃ and extracting their spatial coordinates.

[0108] Perform time-dimensional verification on the extracted red blocks: if the same block shows a red mark in a specified number of consecutive thermal field acquisitions (e.g., 3 times), it is determined to be a heat dissipation blockage area.

[0109] The coordinates of the blocks that have passed continuous verification are aggregated to form a coordinate set {(x1,y1),(x2,y2),…}, which yields the heat dissipation blockage area marker.

[0110] S25. Map the heat dissipation blockage area marker to the corresponding position in the string thermal field and output the coordinates of the local temperature rise anomaly area.

[0111] Specifically, a spatial grid matching algorithm can be used to adapt the coordinate set of the heat dissipation blockage area to the spatial grid system corresponding to the string thermal field, and establish the positional relationship between the two.

[0112] Extract the temperature values ​​T(x,y) of the coordinate matching region and calculate the average temperature T of the same string. avg If the temperature in this region satisfies T(x,y)>T avg If the temperature rise is greater than or equal to ΔT, it is considered an abnormal temperature rise region. Here, ΔT represents the temperature buffer threshold, which can be set to 3-5℃.

[0113] Summarize the coordinates of the abnormal temperature rise areas to form an abnormal coordinate set {(x1,y1)|T(x,y)>T} avg +ΔT} gives the coordinates of the local temperature rise anomaly area.

[0114] In some embodiments, S23, converting the tilt gradient data matrix into a spatial distribution map and outputting the tilt gradient distribution map, includes:

[0115] S231. Perform spatial interpolation on the tilt gradient data matrix to fill in the gap data between components and generate a continuous gradient field.

[0116] Identify the coordinates of missing component gaps in the tilt gradient data matrix, calculate interpolation for each missing location, such as using inverse distance weighted interpolation to calculate the fill value, fill the original matrix with the interpolation results, output complete gridded data, and obtain a continuous gradient field.

[0117] S232. Based on the continuous gradient field, perform pseudo-color mapping to identify the temperature difference intensity and obtain the tilt gradient distribution map.

[0118] The interpolated gradient values ​​are mapped to pseudo-color levels to visualize the risk level.

[0119] For example, the mapping rule could be:

[0120] When θ < 3℃, it is mapped to blue, corresponding to a low-risk area; when 3℃ ≤ θ < 5℃, it is mapped to yellow, corresponding to a medium-risk area; when θ ≥ 5℃, it is mapped to red, corresponding to a high-risk area.

[0121] The location numbers of photovoltaic modules are marked on the image, and the final output is a tilt gradient distribution map in the form of an RGB image matrix.

[0122] In some embodiments, heat loss compensation correction is performed on the coordinates of the local temperature rise anomaly area to analyze the true temperature rise amplitude, including:

[0123] S31. Extract the temperature values ​​of the corresponding components from the coordinates of the local temperature rise anomaly area and output the candidate hot spot temperature set; calculate the theoretical heat loss based on the real-time wind speed and ambient temperature.

[0124] S32. Subtract the theoretical heat loss from the candidate hot spot temperature set and output the corrected temperature value.

[0125] S33. Obtain the average temperature of healthy components in the same string, compare the corrected temperature value with the average temperature, and calculate the actual temperature rise.

[0126] The true temperature rise is obtained by subtracting the average temperature of healthy components from the corrected temperature value.

[0127] In some embodiments, calculating the theoretical heat loss based on real-time wind speed and ambient temperature in S31 includes:

[0128] S311. Calculate the surface convective heat transfer coefficient based on real-time wind speed and ambient temperature, and output the heat transfer coefficient value.

[0129] The heat transfer coefficient h can be calculated using the forced convection heat transfer formula: h = a + b × v cWhere v represents the real-time wind speed, a represents the basic heat transfer coefficient, which is the static convective heat transfer baseline value of the photovoltaic module measured by standard wind tunnel experiments, b represents the wind speed influence coefficient, which is the turbulence correction parameter obtained by fitting the nonlinear relationship between wind speed and heat transfer intensity through wind tunnel experiments of the same standard, characterizing the enhancing effect of wind speed on heat transfer efficiency, and c represents the wind speed exponential correction factor, which is a power exponential parameter optimized based on the measured air flow field data on the surface of the photovoltaic module, used to accurately quantify the contribution of wind speed changes to forced convection heat transfer.

[0130] S312. Calculate the heat loss power based on the heat transfer coefficient value and the standard area of ​​the component, and output the heat loss power value.

[0131] Reference formula for calculating heat loss power: P loss =h×A×(T) surface -T ambient ); where P loss The value represents the heat loss power, A represents the standard area of ​​the photovoltaic module, and T represents the heat loss power value. surface T represents the surface temperature of a photovoltaic module. ambient Represents ambient temperature.

[0132] S313. Convert the heat loss power value into a temperature compensation equivalent and output the theoretical heat loss.

[0133] Reference formula for calculating temperature compensation equivalent: Where, ΔT comp Represents the temperature compensation equivalent, t c Represents the thermal equilibrium time constant, representing the thermal equilibrium time of the module. A default value of 300 seconds can be used. m represents the mass of the photovoltaic module, and C... p This represents the specific heat capacity of silicon.

[0134] The temperature compensation equivalent is the theoretical heat loss.

[0135] In some embodiments, the correlation between the actual temperature rise and real-time power generation data is used to verify and output the faulty component number, including:

[0136] S41. Filter components whose actual temperature rise exceeds the preset temperature rise threshold and generate a list of components to be verified.

[0137] S42. For the list of components to be verified, retrieve the inverter power generation during the same period and output the target power timing curve.

[0138] Extract the power data of the corresponding string of photovoltaic modules from the list, aggregate the power values ​​at specified time intervals (e.g., 15 minutes), generate a time-power curve, and obtain the target power time-series curve.

[0139] S43. Calculate the decay rate of the target power time-series curve relative to the healthy component, and generate the electrothermal correlation factor.

[0140] The formula for calculating the attenuation rate is as follows: Among them, R 衰减 P represents the real-time decay rate. 健康 P represents the power value of a healthy photovoltaic module during the same period. 目标 Represents the power value of the target component during the same time period, max(P) 健康 () represents the maximum power value of a healthy photovoltaic module within the corresponding time period.

[0141] Reference formula for calculating the electrothermal correlation factor: Among them, F 相关 Represents the electrothermal correlation factor, where N represents the number of data sampling points. Represents the real-time attenuation rate of the i-th sampling point. This represents the actual temperature rise during the i-th time period.

[0142] S44. When the electrothermal correlation factor continues to be lower than the preset fault threshold, output the confirmed fault component number.

[0143] The fault threshold can be determined based on historical data analysis of a 10MW photovoltaic power plant. For example, when the electrothermal correlation factor F... 相关 If the fault confirmation rate reaches 90% when the value is less than 0.8, then the fault threshold can be set to 0.8.

[0144] When the electrothermal correlation factor is less than the fault threshold for a specified number of consecutive sampling periods, the photovoltaic module is marked as faulty, and the faulty module number is output.

[0145] In some embodiments, generating a three-dimensional inspection path based on the three-dimensional spatial distribution of the confirmed fault component numbers includes:

[0146] S51. Associate and confirm the three-dimensional geographic coordinates of the faulty component number, and construct a spatial topology map of the fault point.

[0147] Specifically, the three-dimensional geographic coordinates of the photovoltaic module are retrieved by its number. When the distance between adjacent modules is ≤2 meters, a connecting edge is established. The location of the faulty module is used as the node, and the actual physical distance between nodes is used as the edge weight to generate a weighted undirected graph, thus obtaining the spatial topology graph of the fault point.

[0148] S52. Based on the maximum range of the UAV, the spatial topology map is divided to divide the emergency maintenance sub-region.

[0149] Through formula D max =R 续航 ×0.8 Derive the maximum allowable diameter of the subregion, where D maxRepresents the maximum permissible diameter of the spatial distribution of fault points within a sub-region, with a 20% endurance redundancy reserved to cope with flight deviations, R 续航 This represents the drone's nominal maximum range.

[0150] K-means clustering algorithm can be used, with D max Perform partitioning based on constraints:

[0151] The fault point farthest from the fault point in the spatial topology map is selected as the initial cluster center to quickly build a partitioning framework.

[0152] Each faulty node is assigned to the nearest cluster center, while verifying that the spatial diameter of the cluster (the farthest distance between faulty points within the cluster) is ≤ D. max This ensures that a single drone flight can cover the area.

[0153] Repeatedly optimize the cluster center and node allocation until the clustering results are stable, and finally output the emergency maintenance sub-region.

[0154] S53. Generate an obstacle avoidance path based on the parameters of the roof safety passage overlaid on the emergency maintenance sub-area to obtain a three-dimensional inspection path.

[0155] Specifically, for each emergency maintenance sub-region, the fault points are sorted using the Hamilton shortest path algorithm, and the minimum total distance of the fault point access sequence is searched to construct a spatially compact initial inspection path within the sub-region.

[0156] Based on the initial path, roof safety constraints are superimposed: for example, the distance between the path point and the roof edge is not less than a specified distance (e.g., 0.6 meters) to match the fall prevention threshold of the safety passage, and the path is only allowed to traverse areas with a load-bearing pressure less than a specified pressure (e.g., 200 kg / m²). 2 The green-marked load-bearing area is used to avoid structural risks caused by drone loads, and the altitude difference between adjacent path nodes does not exceed a specified degree (such as 15°) to adapt to the drone's climbing performance limit.

[0157] The paths that satisfy the constraints are encapsulated in the format of a region ID and a sequence of path points containing longitude, latitude, and altitude information to obtain a 3D inspection path.

[0158] In some embodiments, a dynamic diagnostic model is output based on the criterion rule for reversing the location of local temperature rise anomaly zones according to maintenance feedback, including:

[0159] S61. Perform maintenance according to the three-dimensional inspection path and obtain the thermal field data after maintenance.

[0160] The command drone flies along a three-dimensional inspection path to the target location and performs maintenance operations such as cleaning or replacement on the photovoltaic module corresponding to the identified faulty component number.

[0161] After maintenance, infrared thermal images are collected at the location of the photovoltaic modules to obtain raw temperature distribution data. After normalization processing, a standardized thermal image is output. Then, real-time wind speed and ambient temperature are fused to generate string thermal field with environmental parameters, thus obtaining the thermal field data after maintenance.

[0162] S62. Compare the post-maintenance thermal field data with the actual temperature rise, and calculate the temperature rise elimination rate.

[0163] S63. Call the tilt angle difference judgment rule used to generate the tilt angle gradient distribution map, adjust the tilt angle change sensitivity parameter according to the temperature rise elimination rate, and output the dynamic diagnostic model.

[0164] Through the formula ΔT adj = k×(1-η)×T th Derive the threshold correction value, where ΔT adj T represents the threshold adjustment amount. th This represents the temperature threshold in the currently effective tilt angle difference determination rule, k represents the adjustment coefficient, and η represents the temperature rise elimination rate.

[0165] The original threshold is added to the adjustment amount to obtain the new threshold T. th_new T th_new =T th +ΔT adj , set the new threshold T th_new Replace T in the original rule th It generates updated tilt angle difference determination logic and outputs dynamic diagnostic model, i.e., updated tilt angle difference determination rules.

[0166] In some embodiments, comparing the post-maintenance thermal field data with the actual temperature rise in S62, and calculating the temperature rise elimination rate, includes:

[0167] S621. Extract the steady-state temperature of the target component from the post-maintenance thermal field data and output the remeasured temperature value.

[0168] Locate and confirm the spatial coordinates corresponding to the faulty component number, and extract the temperature value at the corresponding location from the thermal field data based on these coordinates.

[0169] Wait for the photovoltaic modules to complete the thermal equilibrium process, record the steady-state temperature value at this time, and obtain the retest temperature value corresponding to the faulty module.

[0170] S622. Correlate the remeasured temperature value with the original true temperature rise amplitude at the same location to generate a temperature rise change comparison group.

[0171] Match data from the same date and time period before and after maintenance operations, bind data from the same physical location before and after maintenance by component number, and construct a structured data table in the format of [component number, original temperature rise amplitude, temperature value after maintenance] to obtain a temperature rise change comparison group.

[0172] S623. Calculate the relative attenuation ratio of the temperature rise change comparison group and output the temperature rise elimination rate.

[0173] To calculate the temperature rise reduction, refer to the formula: ΔT 消除 =ΔT 原始 -(T 复测 -T 健康平均 ); where ΔT 消除 Represents the amount of temperature rise eliminated, ΔT 原始 T represents the original true temperature rise. 复测 T represents the retested temperature value. 健康平均 This represents the current average temperature of healthy components in the same string.

[0174] Through the formula: Calculate the relative attenuation ratio, which is the percentage of the original hot spot temperature rise eliminated by the operation and maintenance measures, i.e., the temperature rise elimination rate.

[0175] In some embodiments, this application provides a photovoltaic operation and maintenance management method based on multimodal data, including: acquiring the three-dimensional coordinates of the photovoltaic array and the installation tilt angle data of the photovoltaic modules.

[0176] The drone flies based on the three-dimensional coordinate commands of the photovoltaic array, and simultaneously collects infrared thermal images and environmental parameters to generate string thermal fields with environmental parameters.

[0177] Based on the installation tilt angle data, the tilt angle difference between adjacent components is calculated, a tilt angle gradient distribution map is generated, and the coordinates of the local temperature rise anomaly area are located by combining the string thermal field.

[0178] The coordinates of the local temperature rise anomaly area are corrected for heat loss compensation to analyze the true temperature rise amplitude.

[0179] Obtain real-time power generation data, correlate the actual temperature rise with the real-time power generation data, and verify and output the fault component number.

[0180] A three-dimensional inspection path is generated based on the three-dimensional spatial distribution of the confirmed fault component numbers.

[0181] The command drone executes the three-dimensional inspection path and obtains maintenance feedback. Based on the maintenance feedback, the criteria rule for locating the coordinates of the local temperature rise anomaly area is corrected in reverse, and a dynamic diagnostic model is output.

[0182] In some embodiments, this application provides a photovoltaic operation and maintenance management device based on multimodal data, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program to implement the steps of the photovoltaic operation and maintenance management method based on multimodal data.

[0183] In some embodiments, this application provides a storage medium storing computer program instructions, which are read and executed by a processor to perform the steps of a photovoltaic operation and maintenance management method based on multimodal data.

[0184] In some embodiments, this application provides a computer program product, including a computer program or instructions, wherein when the computer program or instructions are executed by a processor, the steps of a photovoltaic operation and maintenance management method based on multimodal data are implemented.

[0185] Any references to memory, storage, database, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.

[0186] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0187] Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A photovoltaic operation and maintenance management system based on multimodal data, characterized in that, include: Basic data module: Acquires the three-dimensional coordinates of the photovoltaic array and the installation tilt angle data of the photovoltaic modules; Thermal field acquisition module: Based on the three-dimensional coordinate command of the photovoltaic array, the UAV flies and simultaneously acquires infrared thermal images and environmental parameters to generate string thermal fields with environmental parameters; Anomaly location module: Calculates the tilt angle difference between adjacent components based on installation tilt angle data, generates a tilt angle gradient distribution map, and locates the coordinates of local temperature rise anomaly areas by combining the string thermal field; Temperature rise correction module: Performs heat loss compensation correction on the coordinates of the local temperature rise anomaly area and analyzes the true temperature rise amplitude; Fault verification module: acquires real-time power generation data, correlates the actual temperature rise with the real-time power generation data, and verifies and outputs the fault component number; Path planning module: Generates a three-dimensional inspection path based on the three-dimensional spatial distribution of the confirmed fault component numbers; Path planning module: instructs the UAV to execute the three-dimensional inspection path and obtain maintenance feedback, and corrects the judgment rules for locating the coordinates of the local temperature rise anomaly area based on the maintenance feedback, and outputs a dynamic diagnostic model.

2. The method for dynamic monitoring of volatile components in health-promoting plants according to claim 1, characterized in that, The drone flies based on the three-dimensional coordinate commands of the photovoltaic array, simultaneously acquiring infrared thermal images and environmental parameters to generate string thermal fields with environmental parameters, including: Plan the UAV flight path based on the three-dimensional coordinates of the photovoltaic array; The command instructs the drone to fly along the trajectory and acquires the drone's real-time spatial coordinates; The infrared thermal imager is triggered at the real-time spatial coordinates to collect raw temperature distribution data. Normalize the raw temperature distribution data and output a standardized heatmap; It integrates standardized thermal maps with real-time wind speed and ambient temperature from weather stations to output string thermal fields with environmental parameters.

3. The photovoltaic operation and maintenance management system based on multimodal data according to claim 1, characterized in that, Based on the installation tilt angle data, the tilt angle difference between adjacent components is calculated to generate a tilt angle gradient distribution map. Combined with the string thermal field, the coordinates of local temperature rise anomaly areas are located, including: Extract the tilt angle values ​​of adjacent components from the installation tilt angle data; Calculate the absolute value of the difference between adjacent dip angle values ​​to generate a dip angle gradient data matrix; Convert the tilt gradient data matrix into a spatial distribution map and output the tilt gradient distribution map; Identify the coordinates of blocks in the tilt gradient distribution map where the difference continuously exceeds a preset temperature threshold, and mark the heat dissipation blockage area. Map the heat dissipation blockage area markers to the corresponding positions in the string thermal field, and output the coordinates of the local temperature rise anomaly area.

4. The photovoltaic operation and maintenance management system based on multimodal data according to claim 1, characterized in that, The coordinates of the local temperature rise anomaly area are corrected for heat loss compensation, and the true temperature rise amplitude is analyzed, including: Extract the temperature values ​​of the corresponding components from the coordinates of the local temperature rise anomaly area, and output the candidate hot spot temperature set; calculate the theoretical heat loss based on real-time wind speed and ambient temperature. Subtract the theoretical heat loss from the candidate hotspot temperature set and output the corrected temperature value; Obtain the average temperature of healthy components in the same string, compare the corrected temperature value with the average temperature, and calculate the actual temperature rise.

5. The photovoltaic operation and maintenance management system based on multimodal data according to claim 1, characterized in that, By correlating the actual temperature rise with the real-time power generation data, the faulty component number is verified and confirmed, including: Components whose actual temperature rise exceeds the preset temperature rise threshold are selected, and a list of components to be verified is generated. For the list of components to be verified, retrieve the inverter power generation during the same period and output the target power time-series curve; Calculate the decay rate of the target power time-series curve relative to the healthy component, and generate an electrothermal correlation factor; When the electrothermal correlation factor remains below the preset fault threshold, the fault component number is output to confirm the fault.

6. The photovoltaic operation and maintenance management system based on multimodal data according to claim 1, characterized in that, A three-dimensional inspection path is generated based on the three-dimensional spatial distribution of the confirmed fault component numbers, including: By associating and confirming the three-dimensional geographic coordinates of the faulty component number, a spatial topology map of the fault point is constructed. Based on the maximum flight range of the drone, the spatial topology map is segmented to divide the emergency maintenance sub-region; Obstacle avoidance paths are generated by overlaying roof safety passage parameters on the emergency maintenance sub-area, resulting in a three-dimensional inspection path.

7. The photovoltaic operation and maintenance management system based on multimodal data according to claim 3, characterized in that, Based on the criterion rule for reversing the location of local temperature rise anomaly zones through maintenance feedback, a dynamic diagnostic model is output, including: Perform maintenance according to the three-dimensional inspection path and obtain thermal field data after maintenance; Compare the thermal field data after maintenance with the actual temperature rise, and calculate the temperature rise elimination rate. The tilt difference determination rule used to generate the tilt gradient distribution map is invoked, the tilt change sensitivity parameter is adjusted according to the temperature rise elimination rate, and a dynamic diagnostic model is output.

8. The photovoltaic operation and maintenance management system based on multimodal data according to claim 3, characterized in that, Convert the dip gradient data matrix into a spatial distribution map, and output the dip gradient distribution map, including: Spatial interpolation is performed on the tilt gradient data matrix to fill in the component gap data and generate a continuous gradient field; Pseudo-color mapping based on a continuous gradient field is used to identify the temperature difference intensity, resulting in a tilt gradient distribution map.

9. The photovoltaic operation and maintenance management system based on multimodal data according to claim 4, characterized in that, The theoretical heat loss is calculated based on real-time wind speed and ambient temperature, including: The surface convective heat transfer coefficient is calculated based on real-time wind speed and ambient temperature, and the heat transfer coefficient value is output. Calculate the heat loss power based on the heat transfer coefficient value and the standard area of ​​the component, and output the heat loss power value. The heat loss power value is converted into a temperature compensation equivalent, and the theoretical heat loss is output.

10. The photovoltaic operation and maintenance management system based on multimodal data according to claim 7, characterized in that, Compare the post-maintenance thermal field data with the actual temperature rise, and calculate the temperature rise elimination rate, including: Extract the steady-state temperature of the target component from the thermal field data after maintenance, and output the remeasured temperature value; By correlating the remeasured temperature values ​​with the original actual temperature rise at the same location, a temperature rise change comparison group is generated. Calculate the relative attenuation ratio of the temperature rise change comparison group and output the temperature rise elimination rate.

Images