Photovoltaic device operating state monitoring method and system, electronic device and storage medium

Abstract

The application relates to the technical field of solar photovoltaic technology, and particularly provides a photovoltaic device operation state monitoring method and system, an electronic device and a storage medium, wherein the method comprises the following steps: acquiring video image information in a working process of a photovoltaic device and performing pretreatment; detecting the operation state of a photovoltaic panel in the photovoltaic device based on the video image information, and acquiring operation state data of the photovoltaic panel; acquiring posture data of a steel cable and vibration data of the photovoltaic panel in the working process of the photovoltaic device by using a preset sensor detection unit; and determining an abnormality category of the photovoltaic device according to the operation state data, the vibration data and the posture data of the steel cable and outputting display. The purpose is to realize effective and all-round operation state monitoring of the photovoltaic device, to realize real-time control of the operation of multiple parts of the photovoltaic device, and to realize the technical effect of timely discovery of hidden dangers of photovoltaic device faults.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of solar photovoltaic technology, in particular to a photovoltaic device operation state monitoring method and system, an electronic device and a storage medium. BACKGROUND

[0002] At present, in the research of photovoltaic device construction and maintenance, how to accurately monitor the state data information of the photovoltaic device in the operation process has become the key direction of formulating the photovoltaic device fault diagnosis strategy. In the operation process of the flexible photovoltaic support, the photovoltaic panel and the fixed structure such as the steel cable of the flexible photovoltaic support on the flexible photovoltaic support are deformed by long-time shaking due to the influence of the natural environment, which causes the inclination of the flexible photovoltaic support to change, and further affects the power generation efficiency of the photovoltaic device.

[0003] The commonly used photovoltaic device monitoring method in the prior art is usually based on a traditional single sensor detection method, such as using a displacement sensor, a strain sensor, a temperature sensor, etc. to realize monitoring. The existing monitoring method generally detects and analyzes the fault of the photovoltaic panel. Since the monitored items are less and there is a monitoring blind area, the real-time operation condition of the photovoltaic device cannot be comprehensively monitored, and only the monitoring of the photovoltaic panel causes the defects of other parts of the photovoltaic device to be missed. Long-term non-maintenance will cause loss of the photovoltaic device and increase the maintenance cost. Therefore, a new photovoltaic device operation state monitoring method needs to be developed to better realize the state monitoring of the photovoltaic device. SUMMARY

[0004] The purpose of the present application is to provide a photovoltaic device operation state monitoring method, system, electronic device and storage medium, which realizes effective and comprehensive operation state monitoring of the photovoltaic device, real-time control of the operation of multiple parts of the photovoltaic device, and timely discovery of the technical effect of photovoltaic device fault hidden danger.

[0005] The technical solution of the first aspect of the present application provides a photovoltaic device operation state monitoring method, comprising:

[0006] Obtaining video image information in the working process of the photovoltaic device and performing preprocessing;

[0007] Detecting the operation state of the photovoltaic panel in the photovoltaic device based on the video image information, and obtaining the operation state data of the photovoltaic panel;

[0008] Obtaining the posture data of the steel cable and the vibration data of the photovoltaic panel in the working process of the photovoltaic device by using a preset sensor detection unit;

[0009] Determine the abnormal category of the photovoltaic device according to the operation state data, vibration data and posture data of the steel cable of the photovoltaic panel and output display.

[0010] Further, the operation state of the photovoltaic panel in the photovoltaic device is detected based on the video image information, and operation state data of the photovoltaic panel is acquired, specifically comprising:

[0011] The edge information of the photovoltaic panel is extracted by using an edge detection algorithm;

[0012] The extracted edge information is subjected to straight line fitting to obtain an edge segment of the photovoltaic panel;

[0013] The coordinate values of two end points of each edge segment are detected and extracted based on a Hough transform;

[0014] The slope of the edge segment is calculated according to the coordinate values, and then the inclination angle of the photovoltaic panel is calculated, and the calculation formula is:

[0015]

[0016] Wherein, (x1, y1) and (x2, y2) are the coordinates of the two end points of the edge segment, and θ is the included angle between the edge segment and the horizontal line;

[0017] The inclination angle change information of the photovoltaic panel between different time points is recognized by using a matching algorithm.

[0018] The beneficial effects of the above further technical solutions are: the inclination angle of the photovoltaic panel is accurately calculated, the extraction precision and speed of the inclination angle of the photovoltaic panel are improved, and the inclination angle is not affected by factors such as terrain and environment, thereby providing reference data of the inclination angle for operation and maintenance management and fault diagnosis of the photovoltaic panel.

[0019] Further, the operation state of the photovoltaic panel in the photovoltaic device is detected based on the video image information, and operation state data of the photovoltaic panel is acquired, further comprising:

[0020] The pre-processed video image information is subjected to target monitoring based on a pre-constructed target detection module, and the position information of the dark area, cracks and color change of the photovoltaic panel is acquired and labeled.

[0021] The beneficial effects of the above further technical solutions are: the dark area, cracks and color change defects of the cell in the image are automatically detected and labeled, the reliability of the detection result of the photovoltaic panel is improved, and the problem of insufficient monitoring of the cell is solved.

[0022] Further, the construction process of the target detection module comprises:

[0023] Image data sets containing the dark area, cracks and color change of the photovoltaic panel are collected, and the target area is manually labeled;

[0024] Based on a deep learning framework, a Faster R-CNN model is used to train the labeled data set.

[0025] The beneficial effects of the above further technical solutions are that, compared with traditional manual identification and judgment, the target monitoring module can automatically detect the dark areas, cracks and color change defects on a large number of photovoltaic panel cells, and has better accuracy and efficiency.

[0026] Further, the obtaining of the attitude data of the steel cable and the vibration data of the photovoltaic panel during the working process of the photovoltaic device by the preset sensor detection unit comprises:

[0027] obtaining the vibration data information of the photovoltaic panel during working based on the first sensor detection unit;

[0028] obtaining the pressure data information and the displacement data information of the steel cable during working based on the second sensor detection unit;

[0029] obtaining time data corresponding to the vibration data information, the pressure data information and the displacement data information;

[0030] preprocessing the vibration data information, the pressure data information and the displacement data information based on the same time data;

[0031] combining the vibration data, the pressure data and the displacement data based on a multi-sensor data fusion algorithm to obtain multi-source data information.

[0032] The beneficial effects of the above further technical solutions are that, by obtaining different types of data information of the photovoltaic device at the same time, based on the multi-sensor data fusion algorithm, important correlation features between pressure, vibration and displacement signals can be extracted, and they can be combined into more accurate multi-source data results, effectively avoiding misjudgment, and the prediction accuracy of faults is higher.

[0033] Further, the method further comprises: obtaining environmental data information and performing fault analysis and early warning according to the environmental data information and the multi-source data information.

[0034] The technical solution of the second aspect of the present application provides a photovoltaic device running state monitoring system, which comprises:

[0035] an image acquisition module configured to acquire video image information of the photovoltaic device;

[0036] a sensor detection unit configured to obtain attitude data of the steel cable and vibration data of the photovoltaic panel during working of the photovoltaic device;

[0037] a data acquisition module configured to obtain running state data of the photovoltaic panel;

[0038] a processing module configured to perform Gaussian filtering processing on the video image information and normalization processing on the data collected by the sensor detection unit;

[0039] an edge detection module configured to perform edge detection on the photovoltaic panel and calculate a tilt angle of the photovoltaic panel;

[0040] a target detection module configured to detect dark areas, cracks and color changes of the cell pieces on the photovoltaic panel;

[0041] a data fusion module configured to fuse the data collected by the sensor detection unit and the operation state data to form multi-source data information;

[0042] a main control unit configured to read the multi-source data information and perform fault analysis and processing;

[0043] a host computer configured to send a warning signal based on the fault analysis and processing result of the main control unit.

[0044] Further, the sensor detection unit includes a first sensor detection unit and a second sensor detection unit, the first sensor detection unit includes acceleration sensors installed around the bottom of the photovoltaic panel, the second sensor detection unit includes displacement sensors installed on the horizontal plane and the vertical plane of the steel cable, and the second sensor detection unit further includes pressure sensors installed at the end of the steel cable.

[0045] The technical solution of the third aspect of the application provides an electronic device, which includes a processor and a memory connected with the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor to enable the processor to execute the photovoltaic device operation state monitoring method of any one of the technical solutions of the first aspect of the application.

[0046] The technical solution of the fourth aspect of the application provides a computer readable storage medium, which stores a program for implementing the photovoltaic device operation state monitoring method, and the program for implementing the safety detection method is executed by a processor to implement the steps of the photovoltaic device operation state monitoring method of any one of the technical solutions of the first aspect of the application.

[0047] The beneficial effects of the present application include: the present application detects the running state of the photovoltaic panel in the photovoltaic device based on the video image information in the working process of the photovoltaic device, and obtains the running state data of the photovoltaic panel; the posture data of the steel cable and the vibration data of the photovoltaic panel in the working process of the photovoltaic device are obtained by using the preset sensor detection unit; the multi-source information of the photovoltaic panel and the steel cable in the photovoltaic device is collected, and the running state information of the photovoltaic panel is detected based on image recognition, which is used as the basis for judging the abnormal failure of the photovoltaic device, so as to maximize the accuracy of the judgment result; compared with the single monitoring mode of the photovoltaic panel in the prior art, the use of multi-source information and image recognition technology effectively improves the precision of the abnormal failure detection of the photovoltaic device, and more timely detects the failure defect problem, and ensures the safe and stable operation of the photovoltaic power generation equipment. BRIEF DESCRIPTION OF DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation to the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0049] Figure 1 The flowchart of the photovoltaic device running state monitoring method provided by the embodiment of the present application is shown in the figure.

[0050] Figure 2 The structure diagram of the photovoltaic device running state monitoring system provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.

[0052] It should be noted that: similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present application, the terms "first", "second" and the like are only used for distinguishing description, and cannot be understood as indicating or implying relative importance.

[0053] Please refer to Figure 1 The technical scheme of the first aspect of the present application provides a photovoltaic device running state monitoring method, which comprises:

[0054] Step S100: obtaining video image information in the working process of the photovoltaic device and preprocessing;

[0055] In step S100, before video image information of the photovoltaic device is collected, a suitable high-speed camera is selected in advance, the camera is arranged on the photovoltaic device, and it is ensured that the camera can completely capture image information of the photovoltaic panel; in order to guarantee the image quality of the video image information, the pre-processing process of the video image information at least includes: reading the video image information and obtaining the image of the photovoltaic panel according to a preset frame number; converting the image into a gray-scale image; and performing Gaussian filtering on the gray-scale image to reduce noise and interference;

[0056] Step S200: detecting the running state of the photovoltaic panel in the photovoltaic device based on the video image information and obtaining running state data of the photovoltaic panel;

[0057] Step S200 specifically includes:

[0058] Step S210: extracting boundary information of the photovoltaic panel by using an edge detection algorithm; in step S210, the Canny edge detection algorithm is used to extract the edge of the photovoltaic panel in the video image information, and the edge is subjected to binarization processing and morphological operation, and then the contour of the photovoltaic panel is found and the boundary is drawn;

[0059] Step S211: performing straight line fitting on the extracted boundary information to obtain an edge line segment of the photovoltaic panel;

[0060] Step S212: detecting and extracting coordinate values of two end points of each edge line segment based on Hough transformation;

[0061] Step S213: calculating the slope of the edge line segment according to the coordinate values and then calculating the inclination angle of the photovoltaic panel, and the calculation formula is:

[0062]

[0063] wherein (x1, y1) and (x2, y2) are the coordinates of the two end points of the edge line segment, and θ is the included angle between the edge line segment and the horizontal line;

[0064] It should be noted that if the above calculation result needs to be converted into an angle value, the calculation result of the above calculation formula is multiplied by 180 / π; in actual application, a graph of the inclination angle and the vibration state of the photovoltaic panel and data statistical results of each time period can also be generated, thereby providing important reference data basis for operation and maintenance management and fault diagnosis of the photovoltaic panel;

[0065] Step S214: identifying the inclination angle change information of the photovoltaic panel between different time points by using a matching algorithm.

[0066] In the technical solution, the edge detection algorithm based on the video image extracts the boundary information of the photovoltaic panel, and the tilt angle of the photovoltaic panel is calculated by linear fitting and extracting the end point coordinates of the edge line segment. In the prior art, the tilt angle of the photovoltaic panel is usually determined by measuring the angle between the sensor and the ground. The existing method has defects in the measurement result and efficiency of the tilt angle due to the influence of the terrain and environment. The method provided by the embodiment is a non-contact method for determining the tilt angle of the photovoltaic panel. The tilt angle of the photovoltaic panel is calculated based on the computer vision technology, thereby improving the extraction accuracy and speed of the tilt angle of the photovoltaic panel. In actual application, the tilt angle of the photovoltaic panel can be monitored in real time without being affected by the terrain and environment.

[0067] The step S200 further includes:

[0068] Step S220: performing target monitoring on the preprocessed video image information based on a pre-constructed target detection module to obtain position information of dark areas, cracks and color changes of the photovoltaic panel and mark the position information; the dark areas, cracks and color changes of the photovoltaic panel have different degrees of influence on the power generation efficiency and service life of the photovoltaic panel, and therefore these factors of the photovoltaic panel need to be monitored. The existing photovoltaic panel detection method is usually artificial judgment, which has defects of low efficiency and insufficient accuracy. In addition, the existing technology lacks monitoring of the photovoltaic panel. Therefore, the technical solution provided by the embodiment can avoid human interference by using the pre-constructed target detection module to detect the video image. The video image detection method can simultaneously realize automatic monitoring of multiple photovoltaic devices, automatically detect and mark the defects of the dark areas, cracks and color changes of the photovoltaic panel in the image, and improve the reliability of the photovoltaic panel detection result.

[0069] In step S220, the construction process of the target detection module includes:

[0070] Step S221: collecting image data sets containing dark areas, cracks and color changes of the photovoltaic panel, and manually marking the target area;

[0071] The dark areas of the photovoltaic panel usually refer to the areas that cannot effectively convert sunlight into electrical energy. The dark areas will reduce the conversion efficiency of the photovoltaic panel and thus reduce the output power of the entire photovoltaic system. The cracks of the photovoltaic panel may expand and cause problems such as oxidation and corrosion. The existence of the cracks will affect the strength and stability of the photovoltaic panel and accelerate the further expansion of the cracks during long-term use in the external environment. The color change on the surface of the photovoltaic panel mainly refers to the change from deep to shallow or other changes, which is usually caused by oxidation and aging of the photovoltaic panel.

[0072] In step S221, the image dataset should be divided into a training set and a dataset according to a preset ratio. Before manually labeling the image dataset, the image dataset needs to be preprocessed to improve the accuracy of labeling. Existing labeling tools such as Labelbox and CVAT can be used for labeling. During the labeling process, since there may be multiple defects on the photovoltaic panel cell at the same time, different labels need to be used to label the dark area, crack and color change according to the actual situation.

[0073] Step S222: based on the deep learning framework, the Faster R-CNN model is used to train the labeled dataset. After training, the performance of the trained Faster R-CNN model needs to be verified using the test set. The model is optimized and adjusted through evaluation indicators such as recall rate, precision and F1-score.

[0074] In the above technical solution, based on the deep learning framework, the Faster R-CNN model is used to train the labeled dataset. Compared with traditional manual identification and judgment, the target monitoring module can automatically detect a large number of dark areas, cracks and color changes on the photovoltaic panel cell. It has better accuracy and efficiency, which can effectively improve the service life of the photovoltaic device.

[0075] Step S300: using a preset sensor detection unit to obtain the attitude data of the steel cable and the vibration data of the photovoltaic panel during the operation of the photovoltaic device;

[0076] Step S300 specifically includes:

[0077] Step S310: based on the first sensor detection unit, the vibration data information of the photovoltaic panel during operation is obtained; wherein the vibration data of the photovoltaic panel can be used to judge the fault category of the photovoltaic panel; in the above embodiment, the target monitoring model can detect the crack on the photovoltaic panel. If the photovoltaic panel has a crack, the photovoltaic panel vibration data will be abnormal. Therefore, the vibration data can be used to realize the fault warning of the photovoltaic panel.

[0078] Step S320: based on the second sensor detection unit, the pressure data information and displacement data information of the steel cable during operation are obtained; wherein the steel cable of the photovoltaic device may be stretched, worn, deteriorated and other situations during operation. These situations will cause changes in the pressure of the steel cable. Therefore, by monitoring and analyzing the pressure data of the steel cable in real time, the abnormal situation of the steel cable can be detected and located in time, and the potential accident risk can be warned. By collecting and counting the pressure data of the steel cable of the photovoltaic device within a certain period, reference opinions can be provided for the selection of the steel cable material of the photovoltaic device, the layout scheme and the maintenance plan, so as to optimize the photovoltaic device;

[0079] Step S330: acquire time data corresponding to the vibration data information, the pressure data information, and the displacement data information;

[0080] Step S340: pre-process the vibration data information, the pressure data information, and the displacement data information based on the time data; to ensure sufficient reliability of the data, the pre-processing process includes denoising processing, filtering, coordinate system conversion, data normalization, and other pre-processing operations;

[0081] Step S350: combine the vibration data, the pressure data, and the displacement data based on a multi-sensor data fusion algorithm to obtain multi-source data information; wherein, step S350 further includes: calculating corresponding scores or weights according to the consistency between the three data signal elements, thereby screening and combining into more accurate multi-source data results, and realizing accurate evaluation and fault warning of the operation state of the photovoltaic device;

[0082] In the above technical solution, by acquiring different types of data information of the photovoltaic device at the same time, based on the multi-sensor data fusion algorithm, important correlation features between the pressure, vibration, and displacement signals can be extracted, and combined into more accurate multi-source data results, effectively avoiding misjudgment, and then combined with the target detection model and the inclination angle data in the above embodiments, comprehensive evaluation of the operation state of the photovoltaic device can be realized, and the prediction accuracy of the fault is higher;

[0083] Step S400: determine the abnormal category of the photovoltaic device according to the photovoltaic panel operation state data, the vibration data, and the steel cable posture data and output display;

[0084] In step S400, the abnormal category of the photovoltaic device at least includes: output and fault warning of the steel cable pressure value, the steel cable vibration signal, the photovoltaic panel crack, the photovoltaic panel vibration signal, the photovoltaic panel dark area, the photovoltaic panel color, the photovoltaic panel inclination angle, and environmental abnormal signals;

[0085] It should be noted that in the monitoring process of the operation state of the photovoltaic device, environmental data information can also be acquired, and fault analysis and warning can be performed according to the environmental data information and the multi-source data information; for example, temperature, humidity, wind speed, air pressure, and other environmental data can be collected, the collected environmental data and the multi-source data of the photovoltaic device are processed and analyzed, and characteristic parameters are extracted, on this basis, using existing anomaly detection algorithms and correlation analysis algorithms, the potential fault of the multi-source data information under the influence of environmental factors can be detected and warned;

[0086] In summary, the photovoltaic device operation state monitoring method provided by the application detects the operation state of the photovoltaic panel in the photovoltaic device based on video image information in the working process of the photovoltaic device, and obtains operation state data of the photovoltaic panel; the preset sensor detection unit is used to obtain posture data of the steel cable and vibration data of the photovoltaic panel in the working process of the photovoltaic device; the multi-source information of the photovoltaic panel and the steel cable in the photovoltaic device is collected, and the image recognition-based detection of the operation state information of the photovoltaic panel is used as the basis for judging the abnormal failure of the photovoltaic device, so that the accuracy of the judgment result is maximized; compared with the single monitoring mode of the photovoltaic panel in the prior art, the use of multi-source information and image recognition technology effectively improves the accuracy of the abnormal failure detection of the photovoltaic device, and more timely detection of failure defects is achieved, thereby ensuring the safe and stable operation of the photovoltaic power generation equipment.

[0087] Please refer to Figure 2 The technical scheme of the second aspect of the application provides a photovoltaic device operation state monitoring system, which comprises:

[0088] An image acquisition module configured to acquire video image information of the photovoltaic device;

[0089] A sensor detection unit configured to obtain posture data of the steel cable and vibration data of the photovoltaic panel during the working of the photovoltaic device;

[0090] A data acquisition module configured to obtain operation state data of the photovoltaic panel;

[0091] A processing module configured to perform Gaussian filtering on the video image information and normalize the data collected by the sensor detection unit;

[0092] An edge detection module configured to perform edge detection on the photovoltaic panel and calculate the inclination angle of the photovoltaic panel;

[0093] A target detection module configured to detect dark areas, cracks and color changes of the battery pieces on the photovoltaic panel;

[0094] A data fusion module configured to fuse the data collected by the sensor detection unit with the operation state data to form multi-source data information;

[0095] A master control unit configured to read the multi-source data information and perform failure analysis and processing;

[0096] A host computer configured to issue a warning signal based on the failure analysis and processing result of the master control unit.

[0097] Further, the sensor detection unit comprises a first sensor detection unit and a second sensor detection unit, the first sensor detection unit comprises acceleration sensors installed around the bottom of the photovoltaic panel; the second sensor detection unit comprises displacement sensors installed on the horizontal plane and vertical plane of the steel cable; the second sensor detection unit further comprises pressure sensors installed on the end of the steel cable.

[0098] The technical scheme of the third aspect of the present application provides an electronic device, the electronic device comprises a processor and a memory connected with the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor to enable the processor to execute the photovoltaic device running state monitoring method in any one of the technical schemes of the first aspect of the present application.

[0099] The technical scheme of the fourth aspect of the present application provides a computer readable storage medium, the computer readable storage medium stores a program for implementing the photovoltaic device running state monitoring method, and the program for implementing the safety detection method is executed by the processor to implement the steps of the photovoltaic device running state monitoring method in any one of the technical schemes of the first aspect of the present application.

[0100] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for monitoring the operating state of a photovoltaic device, characterized in that The method comprises the following steps: acquiring video image information in the working process of the photovoltaic device and performing preprocessing; detecting the running state of the photovoltaic panel in the photovoltaic device based on the video image information, and acquiring running state data of the photovoltaic panel, which comprises the following steps: extracting boundary information of the photovoltaic panel by using an edge detection algorithm; performing straight line fitting on the extracted boundary information to obtain edge line segments of the photovoltaic panel; detecting and extracting coordinate values of two end points of each edge line segment based on Hough transformation; calculating the inclination angle of the photovoltaic panel according to the coordinate values after calculating the slope of the edge line segment, and the calculation formula is: ； where (x1, y1) and (x2, y2) are the coordinates of the two end points of the edge line segment, is the angle between the edge line segment and the horizontal line; recognizing the inclination angle change information of the photovoltaic panel between different time points by using a matching algorithm; acquiring posture data of the steel cable and vibration data of the photovoltaic panel in the working process of the photovoltaic device by using a preset sensor detection unit, which comprises the following steps: acquiring vibration data information of the photovoltaic panel in the working process based on a first sensor detection unit; acquiring pressure data information and displacement data information of the steel cable in the working process based on a second sensor detection unit; acquiring time data corresponding to the vibration data information, the pressure data information and the displacement data information; performing preprocessing on the vibration data information, the pressure data information and the displacement data information based on the same time data; combining the vibration data, the pressure data and the displacement data based on a multi-sensor data fusion algorithm to obtain multi-source data information; determining the abnormal category of the photovoltaic device according to the running state data of the photovoltaic panel, the vibration data and the posture data of the steel cable and outputting display.

2. The photovoltaic device operating state monitoring method according to claim 1, wherein The method further comprises the following steps: performing target monitoring on the preprocessed video image information based on a pre-constructed target detection module, acquiring position information of dark areas, cracks and color changes of the photovoltaic panel and labeling.

3. The method of claim 2, wherein the method further comprises: The construction process of the target detection module comprises the following steps: collecting image data sets containing dark areas, cracks and color changes of the photovoltaic panel and manually labeling target areas; training the labeled data sets by using a Faster R-CNN model based on a deep learning framework.

4. The photovoltaic device operating state monitoring method according to claim 1, characterized by, The method further comprises the following steps: acquiring environmental data information and performing fault analysis and early warning according to the environmental data information and the multi-source data information.

5. A photovoltaic device operating condition monitoring system, characterized by, The system is used for executing the photovoltaic device running state monitoring method in any one of claims 1 to 4, and comprises the following components: an image acquisition module configured to acquire video image information of the photovoltaic device; a sensor detection unit configured to acquire posture data of the steel cable and vibration data of the photovoltaic panel in the working process of the photovoltaic device; a data acquisition module configured to acquire running state data of the photovoltaic panel; a processing module configured to perform Gaussian filtering on the video image information and to perform normalization on the data acquired by the sensor detection unit; an edge detection module configured to perform edge detection on the photovoltaic panel and to calculate the inclination angle of the photovoltaic panel; a target detection module configured to detect dark areas, cracks and color changes of the photovoltaic panel; a data fusion module configured to fuse the data acquired by the sensor detection unit and the running state data to form multi-source data information. The master control unit is configured to read multi-source data information and perform fault analysis processing. The host computer is configured to issue a warning signal based on the fault analysis processing result of the master control unit.

6. The photovoltaic device operating state monitoring system according to claim 5, wherein The sensor detection unit includes a first sensor detection unit and a second sensor detection unit, the first sensor detection unit includes an acceleration sensor installed around the bottom of the photovoltaic panel; the second sensor detection unit includes a displacement sensor installed on the horizontal plane and the vertical plane of the steel cable; the second sensor detection unit further includes a pressure sensor installed at the end of the steel cable.

7. An electronic device, comprising: The electronic device includes a processor and a memory connected with the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor to enable the processor to execute the photovoltaic device running state monitoring method in any one of claims 1 to 4.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a program for implementing the photovoltaic device running state monitoring method, and the program is executed by the processor to implement the steps of the photovoltaic device running state monitoring method in any one of claims 1 to 4.

Images