A distributed photovoltaic power station adaptive mobile monitoring system

Abstract

The present application relates to photovoltaic power generation technical field, especially relates to a kind of distributed photovoltaic power station self-adapting mobile monitoring system, including mobile inspection module, data reading analysis module, data analysis module, fault monitoring module, fault analysis module, fault alarm module.The present application is by constructing the data set of electrical parameter and temperature parameter, establishes the mapping relationship of both, comprehensively determines electrical abnormal risk and shielding abnormal risk, overcomes the limitation of single parameter monitoring;Adopt multi-parameter correlation analysis, improve fault judgment accuracy;Meanwhile, the present application is by the fusion analysis of picture image and thermal imaging image, utilize threshold segmentation, edge detection and foreign matter feature library matching, accurately identify the type of shielding object;And through picture image and thermal imaging image, judge photovoltaic panel edge deformation, bulge or recess and other structural defects, provide specific processing basis for operation and maintenance, so that the present application can accurately identify the type of shielding and structural defects, enhance operation and maintenance pertinence.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power generation technology, and in particular to an adaptive mobile monitoring system for distributed photovoltaic power stations. Background Technology

[0002] As the global energy structure shifts towards clean energy, distributed photovoltaic (PV) power generation, as an important form of renewable energy utilization, continues to see growth in both application prospects and installed capacity. Among these, PV panels, as the core component of a PV power generation system, directly impact power generation efficiency and system stability. However, PV panels are susceptible to environmental factors during long-term operation, leading to problems such as degraded electrical performance, localized hot spots, and structural deformation. Failure to detect and address these issues promptly can result in a sharp reduction in power generation or even system failure.

[0003] For monitoring photovoltaic (PV) power plants, existing technologies have proposed relevant solutions. For example, patent document CN118801814A discloses a power generation anomaly early warning system and method for PV power generation equipment. This system uses a space radio frequency partial discharge detection radar to detect partial discharge signals of PV equipment and PV power generation network facilities in urban environments. Then, it combines basic information of PV equipment, PV power generation network information, satellite maps of PV power plants, and big data processing models to process and analyze the partial discharge data. Combined with the processing and monitoring of partial discharge events of facilities other than PV equipment in the PV power generation network, it calculates the degree of impact of the event on the operating status of the PV equipment itself. This technical solution mainly analyzes the abnormal status of PV panels through electrical signals.

[0004] For example, patent document CN119628561A discloses a fault monitoring system for photovoltaic power plants based on a neural network model. This system accurately collects string current and voltage data through a data acquisition module, providing a basis for fault diagnosis. The system uses the collected data for fault diagnosis through an IV analysis module, quickly identifying anomalies. The system constructs a power plant model and performs precise positioning through a digital power plant modeling module, supporting inspection route planning. The system generates UAV flight strategies through an inspection route planning module. The system uses machine vision and thermal imaging technology through a CV analysis module to achieve high-precision automatic defect detection.

[0005] In the aforementioned existing technologies, monitoring systems mostly collect electrical parameters, environmental parameters, or image data, but do not establish a correlation mapping relationship between parameters, resulting in one-sided fault judgment. Furthermore, it is difficult to distinguish whether the problem is caused by electrical connection issues or obstruction based solely on voltage anomalies.

[0006] The shading detection accuracy is insufficient, making it impossible to accurately identify the type of shading object and difficult to distinguish between shading and structural defects in the photovoltaic panel itself, resulting in poor targeted operation and maintenance. In addition, the fault early warning mechanism is imperfect, and minor anomalies may be over-processed, increasing operation and maintenance costs. Summary of the Invention

[0007] The present invention solves the above-mentioned technical problems. The present invention adopts the following technical solution: an adaptive mobile monitoring system for distributed photovoltaic power stations, comprising: a mobile inspection module, which sequentially numbers the photovoltaic panels arranged in an array within the photovoltaic power station and automatically inspects them according to a preset inspection track by a mobile inspection vehicle; the mobile inspection vehicle is equipped with a data interface and a video monitoring terminal.

[0008] The data reading and analysis module collects the electrical and temperature parameters of each photovoltaic panel through the data interface, and generates a dataset of electrical and temperature parameters for each photovoltaic panel.

[0009] The data analysis module analyzes the fault status of the photovoltaic panels based on the mapping relationship of the dataset. The fault status includes electrical abnormality risk and shading abnormality risk. If shading abnormality risk exists, the fault monitoring module is executed. If there is no fault, the monitoring continues to the next photovoltaic panel.

[0010] The fault monitoring module acquires surface images and thermal images of the photovoltaic panel through the video monitoring terminal.

[0011] The fault analysis module preprocesses and locates features in surface and thermal imaging images, analyzes the shading type and parameters of the photovoltaic panel, and determines the risk handling method.

[0012] The fault alarm module generates tiered early warnings and handling suggestions based on electrical anomaly risks and obstruction anomaly risks.

[0013] Furthermore, the mobile inspection vehicle is equipped with a mobile robotic arm, which controls the data interface to connect with the data interface of the photovoltaic panel and drives the video monitoring terminal to move and collect surface images and thermal images of different positions of the photovoltaic panel.

[0014] Furthermore, the electrical parameters of the photovoltaic panel include operating current and open-circuit voltage; the temperature parameters include the operating temperature of each location of the photovoltaic panel cells, and the dataset consists of electrical parameters and temperature parameters at each preset time point.

[0015] Furthermore, the fault state analysis method is as follows: S1, read the operating temperature of each position of the battery cell, and calculate the average operating temperature of each photovoltaic panel by averaging. Filter the highest and lowest operating temperatures of each photovoltaic panel, and take the absolute value of the difference between them and the average operating temperature of each photovoltaic panel to obtain the upper deviation temperature and lower deviation temperature of each photovoltaic panel.

[0016] S2. Calculate the theoretical open-circuit voltage value under the current temperature conditions using the model of the correspondence between photovoltaic panel temperature and open-circuit voltage; compare it with the actual collected open-circuit voltage. If the deviation exceeds the preset threshold, it is determined that the photovoltaic panel has an electrical abnormality risk; otherwise, it does not.

[0017] S3. Based on the changing trend of the operating current of each photovoltaic panel, mark the photovoltaic panels whose operating current is lower than the set operating current threshold, and determine whether there is a risk of abnormal shading by combining the upper deviation temperature and the lower deviation temperature.

[0018] S4. Based on the results obtained from S2 to S3, determine whether the photovoltaic panel is in a faulty state.

[0019] Furthermore, the method for determining the shading abnormality risk of each photovoltaic panel is as follows: if the upper deviation temperature and lower deviation temperature of the photovoltaic panel are greater than the preset deviation temperature threshold, or its operating current is lower than the set operating current threshold, then the photovoltaic panel is determined to have a shading abnormality risk; otherwise, it does not.

[0020] Furthermore, the set operating current threshold is the average operating current of this type of photovoltaic panel during the same period in history multiplied by a proportional coefficient, and the proportional coefficient is greater than 0 and less than 1.

[0021] Furthermore, the risk parameter analysis method for the photovoltaic panel is as follows: read the surface image of the photovoltaic panel, perform noise reduction and contrast enhancement processing on it, and locate the shading area through threshold segmentation; read the thermal imaging image, perform temperature calibration and map the temperature value to pseudo-color, and mark the temperature abnormal area; align the coordinates of the surface image and the thermal imaging image and associate the shading area with the temperature abnormal area to form a spatial-temperature correlation dataset; extract the contour of the shading area through edge detection, calculate the shading parameters and color parameters of the shading area, and combine the spatial-temperature correlation dataset to determine the shading type of the shading area.

[0022] Furthermore, the occlusion parameters include the area, perimeter, shape factor, and texture of the occlusion region; the color parameters are obtained by extracting the hue and saturation of the occluder from the color space of the occlusion region; the average temperature and temperature standard deviation of the corresponding region in the thermal imaging image are extracted, and the occluder is determined by the foreign object feature library, and the processing method is matched according to the type of occluder.

[0023] Furthermore, the foreign object feature database includes a dataset of different types of obstructions in terms of color space and temperature characteristics. By comparing the hue, saturation, and temperature distribution of the obstruction area with the thermal imaging image, the type of obstruction is matched. If the foreign object features are unclear and the structure is abnormal, a second shooting is initiated using a mobile monitoring device. The shooting angle is adjusted to obtain an orthophoto image of the photovoltaic panel surface. At the same time, the thermal imaging sampling frequency is increased to analyze whether the temperature of the obstruction area is stable over time. If the temperature is unstable, it is marked as a foreign object to be updated. After manual confirmation, it is updated to the foreign object feature database.

[0024] Furthermore, the fault analysis module also analyzes whether the photovoltaic panel is deformed by using surface images and thermal imaging images. The specific analysis method is as follows: the straight-line fitting degree of the photovoltaic panel edge in the surface image is detected and analyzed by Hough transform. If the straight-line fitting degree is greater than a set threshold, the edge of the photovoltaic panel is determined to be deformed. The surface height difference of the photovoltaic panel in the surface image is calculated by SIFT feature point matching. If the surface height difference is greater than a set threshold, and the temperature distribution in the thermal imaging image shows an abnormal gradient, and the area is unobstructed, the area of ​​the photovoltaic panel is determined to be bulging or dented.

[0025] The beneficial effects of this system are as follows: First, this invention enables automatic inspection of photovoltaic panels by a mobile inspection vehicle along a preset track, combined with a mobile robotic arm to achieve automatic data interface docking and multi-position image acquisition, replacing traditional manual inspection or drone inspection. At the same time, it achieves high-frequency, full-coverage monitoring, avoids missed inspections, and is especially suitable for large-scale photovoltaic arrays, improving inspection efficiency and comprehensiveness.

[0026] Second, this invention constructs a dataset of electrical parameters and temperature parameters, establishes a mapping relationship between the two, and comprehensively determines the risks of electrical anomalies and shading anomalies, overcoming the limitations of single-parameter monitoring; it also employs multi-parameter correlation analysis to improve the accuracy of fault diagnosis.

[0027] Third, this invention accurately identifies the type of obstruction by fusing surface images and thermal imaging images, using threshold segmentation, edge detection, and foreign object feature library matching; and by using surface images and thermal imaging images to determine structural defects such as edge deformation, bulges, or dents of photovoltaic panels, it provides specific processing basis for operation and maintenance, enabling this invention to accurately identify the type of obstruction and structural defects, and enhance the targeted nature of operation and maintenance.

[0028] Fourth, this invention implements a multi-level alarm mechanism based on fault type, prioritizing the handling of electrical anomalies and setting processing time limits for obstruction anomalies according to their type, thereby improving emergency response efficiency. Simultaneously, by updating the foreign object feature database through secondary imaging and manual confirmation, the system's ability to identify unknown obstructions is enhanced, adapting to complex environments.

[0029] Fifth, the alarm terminal of this invention displays a heat map of the risk location and occlusion parameters in real time. Combined with the precise positioning function, it facilitates maintenance personnel to quickly reach the fault point, shortens the troubleshooting time, and reduces maintenance costs. Attached Figure Description

[0030] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0031] Figure 1 This is a connection logic diagram between the various modules of this invention.

[0032] Figure 2 This is a block diagram of the three-level alarm of the fault alarm module of the present invention. Detailed Implementation

[0033] The embodiments of the present invention are described in detail below. The embodiments described below are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. Where specific techniques or conditions are not specified in the embodiments, they shall be performed in accordance with the techniques or conditions described in the literature in the art or in accordance with the product manual.

[0034] See Figure 1 An adaptive mobile monitoring system for a distributed photovoltaic power station includes: a mobile inspection module, a data reading and analysis module, a data analysis module, a fault monitoring module, a fault analysis module, and a fault alarm module; wherein the mobile inspection module is connected to the data reading and analysis module and the fault monitoring module respectively; the fault analysis module is connected to the data reading and analysis module, the fault alarm module, and the fault monitoring module respectively; and the fault analysis module is connected to the fault monitoring module and the fault alarm module respectively.

[0035] The mobile inspection module sequentially numbers the photovoltaic panels arranged in an array within a photovoltaic power station and automatically inspects them along a preset inspection track using a mobile inspection vehicle. The mobile inspection vehicle is equipped with a data interface and a video monitoring terminal. This invention's mobile inspection module can collect photovoltaic panel data and images through the data interface and video monitoring terminal, respectively.

[0036] It should be noted that the preset inspection track is a fixed guide rail laid between the photovoltaic arrays, and the data interface is a physical interface (such as RS485 or CAN bus interface) used to communicate with the data acquisition unit or smart junction box built into the photovoltaic panel to obtain electrical and temperature parameters. The mobile inspection module includes a mobile inspection vehicle equipped with a mobile robotic arm. The mobile robotic arm controls the data interface to connect with the data interface of the photovoltaic panel and drives the video monitoring terminal to move and collect surface images and thermal images of different positions of the photovoltaic panel.

[0037] The data reading and analysis module collects the electrical and temperature parameters of each photovoltaic panel through the data interface, and generates a dataset of electrical and temperature parameters for each photovoltaic panel.

[0038] The electrical parameters of the photovoltaic panel include the operating current and open-circuit voltage; the temperature parameters include the operating temperature of each location of the photovoltaic cell. The dataset consists of the electrical and temperature parameters at each preset time point.

[0039] In the above dataset, the open-circuit voltage of the photovoltaic panel is the terminal voltage of the photovoltaic panel, which is mainly determined by the material characteristics and temperature; the operating current is the operating current of the photovoltaic panel when it actually generates electricity, which mainly reflects the power generation of the photovoltaic panel; the setting of each preset time point in the dataset can ensure the comprehensiveness of the photovoltaic panel in the data acquisition process and prevent the overall data from being inaccurate due to the deviation of instantaneous data acquisition.

[0040] The data analysis module analyzes the fault status of photovoltaic panels based on the mapping relationship of the dataset. The fault status includes electrical abnormality risk and shading abnormality risk. If shading abnormality risk exists, the fault monitoring module is executed. If there is no fault, the monitoring continues to the next photovoltaic panel.

[0041] The fault status analysis method is as follows: S1. Read the operating temperature of each position of the battery cell and calculate the average operating temperature of each photovoltaic panel. Filter the highest and lowest operating temperatures of each photovoltaic panel, and take the absolute value of the difference between them and the average operating temperature of each photovoltaic panel to obtain the upper deviation temperature and lower deviation temperature of each photovoltaic panel.

[0042] S2. Using a model that correlates photovoltaic panel temperature with open-circuit voltage, calculate the theoretical open-circuit voltage value under the current temperature conditions. Compare this value with the actual collected open-circuit voltage. If the deviation exceeds a preset threshold, the photovoltaic panel is deemed to have an electrical anomaly risk; otherwise, it is not. The logic for judging electrical anomaly risk is that if the open-circuit voltage deviation of the photovoltaic panel is large under specific temperature conditions, there is a problem with the electrical connection.

[0043] S3. Based on the changing trends of the operating current of each photovoltaic panel, mark the photovoltaic panels whose operating current is lower than the set operating current threshold, and determine whether there is an abnormal risk of shading based on the upper and lower deviation temperatures. Since the change in operating current reflects the power generation of the photovoltaic panel, when the power generation of the photovoltaic panel changes abnormally, and the upper or lower deviation temperature of the photovoltaic panel exceeds the set threshold, then the photovoltaic panel is shading-prone.

[0044] For example, when there are metallic foreign objects on the photovoltaic panel, the temperature of the metal will be higher than that of the photovoltaic panel after absorbing heat, which will cause the upper deviation temperature of the photovoltaic panel to exceed the set threshold. When there are shading objects on the photovoltaic panel, these objects have poor heat absorption and temperature rise, and they will block the photovoltaic panel and affect its temperature rise, causing the lower deviation temperature of the photovoltaic panel to exceed the set threshold.

[0045] S4. Combining the results obtained from S2 to S3, determine whether the photovoltaic panel is in a faulty state. This invention can analyze whether the photovoltaic panel has electrical abnormality risks and shading abnormality risks by mapping the electrical parameter-temperature parameter dataset, which facilitates subsequent inspection or maintenance.

[0046] The model relating photovoltaic panel temperature to open-circuit voltage is based on the physical characteristics of photovoltaic cells, and its expression is as follows: .

[0047] in, The open-circuit voltage under standard test conditions. Temperature coefficient of open-circuit voltage for photovoltaic panels (unit: V / ℃), provided by the manufacturer or obtained through experimental calibration, where T is the current measured temperature. The theoretical open-circuit voltage at the current temperature is calculated using the model relating photovoltaic panel temperature to open-circuit voltage, and then compared with the measured value.

[0048] The method for determining the shading anomaly risk of each photovoltaic panel is as follows: if the upper deviation temperature or lower deviation temperature of the photovoltaic panel is greater than the preset deviation temperature threshold, or its operating current is lower than the set operating current threshold, then the photovoltaic panel is determined to have an shading anomaly risk; otherwise, it is not. Large deviations in photovoltaic panel temperature and operating current are obvious signs that the photovoltaic panel is being shaded.

[0049] As one embodiment, the set operating current threshold is the average historical operating current of this type of photovoltaic panel during the same period multiplied by a proportionality coefficient, where the proportionality coefficient is greater than 0 and less than 1. This invention compares the operating current with historical operating currents during the same period, reducing the probability of identifying current anomalies. In a preferred embodiment, the proportionality coefficient is set to 0.83, and this value can be adjusted in the system configuration according to actual operating conditions.

[0050] It should be emphasized that the parameters of the photovoltaic panels are monitored under clear weather conditions in this invention, and therefore the impact of weather on the power generation of the photovoltaic panels is not considered.

[0051] When photovoltaic (PV) panels pose a risk of abnormal shading, further in-depth monitoring is required. This invention employs video monitoring and analysis to acquire image information of PV panels at risk of abnormal shading. The fault monitoring module obtains surface images and thermal images of the PV panel through the video monitoring terminal. The surface images and thermal images are then analyzed to determine the three-dimensional features and thermal distribution of the PV panel surface, facilitating the identification of the morphology of any obstructions on the PV panel.

[0052] The fault analysis module preprocesses and locates features in surface and thermal imaging images, analyzes the shading type and parameters of the photovoltaic panel, and determines the risk handling method.

[0053] The risk types and risk parameter analysis methods for photovoltaic panels are as follows: Surface images of the photovoltaic panels are read, and noise reduction and contrast enhancement processing are performed. Obstruction areas are located through threshold segmentation. The threshold segmentation uses the Otsu algorithm. After contrast enhancement processing, the obstruction areas will show a grayscale difference between the obstruction and the photovoltaic panel. Areas with a grayscale difference greater than 30% between the photovoltaic panel and the background are identified as candidate areas.

[0054] The thermal imaging image is read, temperature is calibrated, and the temperature values ​​are mapped to pseudo-color to mark areas of temperature anomaly. The pseudo-color uses a gradient from blue to red corresponding to -10 to 60℃, and areas with a temperature difference greater than 5℃ are marked as temperature anomaly areas.

[0055] By aligning the coordinates of surface images and thermal images and associating occluded regions with temperature anomaly regions, a spatial-temperature correlation dataset is formed. This dataset links the two regions together, enabling comparison of corresponding locations in the surface and thermal images, thus increasing the intuitiveness of occlusion analysis.

[0056] The contours of occluded regions are extracted through edge detection, and the occlusion and color parameters of these regions are calculated. The occlusion type is then determined by combining this data with a spatial-temperature correlation dataset. The Canny algorithm is used to extract the contours of these occluded regions.

[0057] The occlusion parameters include the area S, perimeter L, shape factor, and texture of the occluded region; the color parameters are obtained by extracting the hue and saturation of the occluded object from the color space of the occluded region; the average temperature and temperature standard deviation of the corresponding region in the thermal imaging image are extracted, and the occluded object is determined by the foreign object feature library, and the processing method is matched according to the type of occluded object.

[0058] For example, the formula for calculating the shape factor of the occluded area is: ,in Represents the shape factor. The value of pi is represented; the roundness of the occluded object is determined by calculating the shape factor. The texture is obtained by calculating the contrast using the gray-level co-occurrence matrix.

[0059] The foreign object feature library is constructed based on historical inspection data and manually labeled samples. Machine learning methods (such as support vector machines or convolutional neural networks) are used to train and classify the image and thermal features of occluded objects. The initial feature library contains multi-dimensional feature data of common occluder types (such as bird droppings, branches, dust, metal fragments, plastic film, etc.). When the system identifies occluders with unknown or ambiguous features during inspections, it automatically triggers a secondary shooting and manual confirmation process. Confirmed samples are added to the feature library and the model is retrained, enabling dynamic updates and optimization of the feature library.

[0060] The foreign object feature library includes a set of data on the color space and temperature characteristics of different types of occluders. By comparing the hue, saturation and temperature distribution of the occluded area and the thermal imaging image, the type of occluder is matched.

[0061] The occlusion parameters and color parameters of common occluders in the foreign object feature library are shown in Table 1 below:

[0062]

[0063] The table above only records parameters of several common obstructions. Due to space limitations, it will not be shown in detail here.

[0064] If the foreign object's characteristics are unclear and its structure is abnormal, meaning the foreign object feature database cannot match the obstruction, a second shooting is initiated using mobile monitoring equipment. The shooting angle is adjusted to obtain an orthophoto of the photovoltaic panel surface, and the thermal imaging sampling frequency is increased to analyze whether the temperature of the obstructed area stabilizes over time. If the temperature is unstable, it is marked as a foreign object awaiting update; after manual confirmation, it is updated to the foreign object feature database; if the temperature is stable, it is not processed. This invention determines various parameters of the obstruction through multi-angle secondary shooting, facilitating the storage of parameters for foreign objects updated in the foreign object feature database.

[0065] The fault analysis module also analyzes whether the photovoltaic panel is deformed using surface images and thermal imaging images. Specifically, it uses Hough transform to detect and analyze the straight-line fit of the photovoltaic panel edges in the surface image. If the straight-line fit is greater than a set threshold, the photovoltaic panel edge is determined to be deformed. For example, this invention uses a straight-line fit greater than 5 pixels to determine if the photovoltaic panel edge is deformed.

[0066] The surface height difference of the photovoltaic panel in the surface image is calculated by SIFT feature point matching. If the surface height difference is greater than a set threshold, and the temperature distribution in the thermal imaging image shows an abnormal gradient, and the area is unobstructed, then the photovoltaic panel is determined to have a bulge or depression in that area. For example, SIFT feature point matching is used to calculate the three-dimensional coordinates of feature points on the photovoltaic panel surface. These three-dimensional coordinates are calculated based on binocular vision or motion reconstruction algorithms. If the local point cloud height difference is >5mm, and the photovoltaic panel in that area is unobstructed and the temperature distribution in the thermal imaging image shows an abnormal gradient, then the photovoltaic panel is determined to have a bulge or depression.

[0067] This invention analyzes whether photovoltaic panels have edge deformation and surface anomalies by using surface images and thermal imaging images, avoiding the need for conventional monitoring to only monitor the covering or structure of the photovoltaic panel. This increases the comprehensiveness of photovoltaic panel monitoring. At the same time, the accuracy of photovoltaic panel-related analysis can be increased by using Hough transform detection, binocular vision or motion recovery structure algorithm.

[0068] See Figure 2 The fault alarm module generates graded early warnings and handling suggestions based on electrical abnormality risks and obstruction abnormality risks. For example, the graded early warning adopts a three-level alarm mechanism, where electrical abnormality risks correspond to the first-level alarm, which has the highest priority; obstruction abnormality risks correspond to the second-level alarm and the third-level alarm.

[0069] When a Level 1 alarm is triggered, the system immediately issues an audible and visual alarm and pushes an emergency notification to the operation and maintenance terminal in real time. The handling suggestions include: immediately disconnecting the electrical connection of the relevant photovoltaic panels, checking for short circuits or open circuits in the cells or lines, and completing on-site repairs within the specified time.

[0070] When a Level 2 alarm is triggered, the system generates a visual alarm and sends a warning report. The handling suggestions are combined with the output of the fault analysis module, including: if it is metal fragments or tree branches, it is recommended to remove them within a time limit to avoid hot spot effect, with a handling time limit of 48 hours; if the type of obstruction is unknown, manual investigation is required within 72 hours.

[0071] If the obstruction is bird droppings or dust, it corresponds to a level three alarm, and it is recommended to arrange for non-emergency cleaning.

[0072] In addition, the alarm terminal integrates a space-temperature correlation dataset, displaying real-time heat maps and shading parameters for risky locations, facilitating rapid identification by maintenance personnel. For normal, risk-free conditions, the system only records monitoring logs and continues patrolling the next photovoltaic panel, ensuring the continuity and efficiency of the monitoring process.

[0073] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention, which are still covered within the protection scope of the present invention.

Claims

1. An adaptive mobile monitoring system for distributed photovoltaic power stations, characterized in that, include: The mobile inspection module sequentially numbers the photovoltaic panels arranged in the array within the photovoltaic power station and performs automatic inspections according to a preset inspection track using a mobile inspection vehicle. The mobile inspection vehicle is equipped with a data interface and a video monitoring terminal. The data reading and analysis module collects the electrical and temperature parameters of each photovoltaic panel through the data interface, and generates an electrical parameter-temperature parameter dataset for each photovoltaic panel. The electrical parameters include the operating current and open-circuit voltage, and the temperature parameters include the operating temperature of each position of the photovoltaic cell. The dataset consists of the electrical and temperature parameters at each preset time point. The data analysis module analyzes the fault status of the photovoltaic panels based on the mapping relationship of the dataset. The fault status includes electrical abnormality risk and shading abnormality risk. If shading abnormality risk exists, the fault monitoring module is executed. If there is no fault, the monitoring continues to the next photovoltaic panel. The specific analysis method for the fault status is as follows: S1. Read the operating temperature of each position of the solar cell and calculate the average operating temperature of each photovoltaic panel. Filter the highest and lowest operating temperatures of each photovoltaic panel, and take the absolute value of the difference between them and the average operating temperature of each photovoltaic panel to obtain the upper deviation temperature and lower deviation temperature of each photovoltaic panel. S2. Calculate the theoretical open-circuit voltage value under the current temperature conditions using the photovoltaic panel temperature and open-circuit voltage correspondence model; compare it with the actual collected open-circuit voltage. If the deviation exceeds the preset threshold, it is determined that the photovoltaic panel has an electrical abnormality risk; otherwise, it does not. S3. Based on the changing trend of the working current of each photovoltaic panel, mark the photovoltaic panels whose working current is lower than the set working current threshold, and determine whether there is a risk of abnormal shading by combining the upper deviation temperature and the lower deviation temperature. S4. Based on the results obtained from S2 to S3, determine whether the photovoltaic panel is in a faulty state. The fault monitoring module acquires surface images and thermal images of the photovoltaic panel through the video monitoring terminal; The fault analysis module preprocesses and locates features in surface and thermal imaging images, analyzes the shading type and shading parameters of photovoltaic panels, and determines the risk handling method. The fault alarm module generates tiered early warnings and handling suggestions based on electrical anomaly risks and obstruction anomaly risks.

2. The adaptive mobile monitoring system for distributed photovoltaic power stations according to claim 1, characterized in that, The mobile inspection vehicle is equipped with a mobile robotic arm. The mobile robotic arm controls the data interface to connect with the data interface of the photovoltaic panel, and drives the video monitoring terminal to move and collect surface images and thermal images of different positions of the photovoltaic panel.

3. The adaptive mobile monitoring system for a distributed photovoltaic power station according to claim 1, characterized in that, The method for determining the abnormal shading risk of each photovoltaic panel is as follows: If the upper or lower deviation temperature of the photovoltaic panel is greater than the preset deviation temperature threshold, or if its operating current is lower than the set operating current threshold, then the photovoltaic panel is determined to have an abnormal shading risk; otherwise, it is not.

4. The adaptive mobile monitoring system for a distributed photovoltaic power station according to claim 1, characterized in that, The set operating current threshold is the average value of the historical operating current of this type of photovoltaic panel during the same period multiplied by a proportional coefficient, and the proportional coefficient is greater than 0 and less than 1.

5. The adaptive mobile monitoring system for a distributed photovoltaic power station according to claim 1, characterized in that, The risk types and risk parameter analysis methods for the photovoltaic panels are as follows: Read the surface image of the photovoltaic panel, perform noise reduction and contrast enhancement processing on it, and locate the shading area by threshold segmentation; Read thermal imaging images, perform temperature calibration, map temperature values ​​to pseudo-color, and mark areas of temperature anomalies; By aligning the coordinates of surface images and thermal imaging images and associating occluded areas with temperature anomaly areas, a spatial-temperature correlated dataset is formed. The contours of occluded regions are extracted by edge detection, the occlusion parameters and color parameters of the occluded regions are calculated, and the occlusion type of the occluded regions is determined by combining the spatial-temperature correlation dataset.

6. The adaptive mobile monitoring system for a distributed photovoltaic power station according to claim 5, characterized in that, The occlusion parameters include the area, perimeter, shape factor, and texture of the occlusion region; the color parameters are obtained by extracting the hue and saturation of the occluder in the color space of the occluded region; the average temperature and temperature standard deviation of the corresponding region in the thermal imaging image are extracted, and the occluder is determined by the foreign object feature library, and the processing method is matched according to the type of occluder.

7. The adaptive mobile monitoring system for a distributed photovoltaic power station according to claim 6, characterized in that, The foreign object feature library includes a data set of different types of occluders in terms of color space and temperature characteristics. By comparing the hue, saturation and temperature distribution of the occluded area and the thermal imaging image, the type of occluder is matched. If the foreign object's characteristics are unclear and its structure is abnormal, the mobile monitoring equipment is activated to take a second picture, and the shooting angle is adjusted to obtain an orthophoto of the photovoltaic panel surface. At the same time, the thermal imaging sampling frequency is increased to analyze whether the temperature of the shaded area is stable over time. If the temperature is unstable, it is marked as a foreign object to be updated. After manual confirmation, it is updated to the foreign object feature database.

8. The adaptive mobile monitoring system for a distributed photovoltaic power station according to claim 5, characterized in that, The fault analysis module is also used to analyze whether the photovoltaic panel is deformed through surface images and thermal imaging images. The specific analysis method is as follows: The straight-line fitting degree of the photovoltaic panel edge in the surface image is detected and analyzed by Hough transform. If the straight-line fitting degree is greater than a set threshold, the photovoltaic panel edge is determined to be deformed. The surface height difference of the photovoltaic panel in the surface image is calculated by SIFT feature point matching. If the surface height difference is greater than a set threshold, and the temperature distribution in the thermal imaging image shows an abnormal gradient, and the area is unobstructed, then the photovoltaic panel is determined to be bulging or dented in that area.

Images