Photovoltaic power station iv, cv fusion diagnosis method and system

Abstract

The application provides a photovoltaic power station IV, CV fusion diagnosis method and system, relates to the technical field of IV, CV, and specifically comprises the following steps: obtaining historical time period operation data and operation state of a photovoltaic power station to be diagnosed, extracting data characteristics, obtaining open-circuit voltage ratio, crack severity index and other parameters, processing related parameters, generating a battery comprehensive index for evaluating battery health and a photovoltaic panel comprehensive index for evaluating photovoltaic panel thermal stability, constructing an operation state recognition model, training the model with the comprehensive index of the historical time period as input and the operation state as label, inputting the comprehensive index of the current time period into the trained model, obtaining the operation state of the current time period, and realizing accurate diagnosis of the operation state of the photovoltaic power station. The application accurately judges the operation state of the photovoltaic power station, and avoids the inaccurate diagnosis problem caused by simply fusing two diagnosis results in the prior art.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of IV and CV, in particular to an IV and CV fusion diagnosis method and system for photovoltaic power stations. BACKGROUND

[0002] With the growing demand for clean energy worldwide, photovoltaic power stations, as a sustainable energy solution, have been widely used and rapidly developed. Photovoltaic power stations convert solar energy into electricity, providing green power for society, which is of great significance for alleviating energy crisis and reducing environmental pollution. However, with the continuous expansion of photovoltaic power station scale and the increase of operation time, its reliability and stability face many challenges. The IV and CV fusion diagnosis method for photovoltaic power stations aims to comprehensively utilize the electrical performance information obtained by IV diagnosis technology and the image information obtained by CV diagnosis technology, and through deep fusion of IV data and CV data, the advantages of the two technologies are fully utilized, and the shortcomings of each other are made up, so as to improve the accuracy and reliability of fault diagnosis.

[0003] In the prior art, the IV&CV fusion diagnosis system with publication number CN117353658A includes the following steps: a fusion diagnosis server, an IV diagnosis server and a drone, the fusion diagnosis server is in communication connection with the IV diagnosis server and the drone, and is used to execute a diagnosis task in response to a diagnosis task request, including: triggering the IV diagnosis server to perform IV diagnosis analysis on current data and voltage data of photovoltaic modules in a target photovoltaic power station, and receiving IV diagnosis analysis results sent by the IV diagnosis server; controlling the drone to fly and collect images of the photovoltaic modules in the target photovoltaic power station to obtain photovoltaic module images, and identifying the photovoltaic module images to obtain CV diagnosis analysis results; matching and fusing the IV diagnosis analysis results and the CV diagnosis analysis results to obtain the final diagnosis analysis results. The IV&CV fusion diagnosis system can improve the fault diagnosis rate of photovoltaic modules by fusing IV and CV to diagnose photovoltaic modules.

[0004] However, there are still the following deficiencies. As can be seen from the above statement, the prior art has the problem of single data dimension, relying only on IV diagnosis analysis results and CV image recognition results, which is difficult to detect heat-related potential faults, leading to possible omission of diagnosis; at the same time, there is a lack of fault training process, directly matching and fusing diagnosis results, which is difficult to accurately judge the fault type in the face of complex and variable fault conditions; and the diagnosis process is not fine enough, simply fusing two diagnosis results without detailed classification and identification of faults.

[0005] The above information disclosed in the background section is only used to enhance the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0006] The application aims to provide a photovoltaic power station IV, CV fusion diagnosis method and system to solve the problems in the background.

[0007] To achieve the above-mentioned purpose, the application provides the following technical solutions:

[0008] A photovoltaic power station IV, CV fusion diagnosis method, the specific steps include:

[0009] S1. Obtain the operation data and operation state of the photovoltaic power station to be diagnosed in the historical time period, the operation data including IV data, image data and infrared thermal imaging data, the operation state including one or more state combinations of normal, hot spot, hidden crack, power attenuation and poor welding contact;

[0010] S2. Feature extraction is performed on the IV data, image data and infrared thermal imaging data to obtain the feature parameters of the photovoltaic power station to be diagnosed in the historical time period, including the open circuit voltage ratio of the photovoltaic cell, the short circuit current ratio, and the crack severity index and temperature deviation index of the photovoltaic panel surface;

[0011] S3. Data processing is performed on the open circuit voltage ratio and short circuit current ratio of the photovoltaic cell to generate a cell comprehensive index for evaluating the health status of the photovoltaic cell, and data processing is performed on the crack severity index and temperature deviation index of the photovoltaic panel surface to generate a photovoltaic panel comprehensive index for evaluating the thermal stability of the photovoltaic panel;

[0012] S4. Constructing an operation state recognition model, taking the cell comprehensive index and the photovoltaic panel comprehensive index of each historical time period of the photovoltaic power station to be diagnosed as input, and taking the operation state as label, training the operation state recognition model;

[0013] S5. Obtaining the cell comprehensive index and the photovoltaic panel comprehensive index of the photovoltaic power station to be diagnosed in the current time period, inputting the trained operation state recognition model, and obtaining the operation state of the photovoltaic power station to be diagnosed in the current time period.

[0014] Further, the length of each time period is set to 1 day, the time interval between adjacent time periods is equal, the open circuit voltage ratio and the short circuit current ratio of the photovoltaic cell are calculated, and the formula is as follows:

[0015]

[0016] Wherein, OVR i is the open circuit voltage ratio of the photovoltaic cell in the i-th historical time period, SCR i is the short circuit current ratio of the photovoltaic cell in the i-th historical time period, is the actual open circuit voltage of the photovoltaic cell in the i-th historical time period, V oc0 is the standard open circuit voltage of the photovoltaic cell, Ii is the actual short-circuit current of the photovoltaic cell for the i-th historical time period, V sc0 Ii is the standard short-circuit current of the photovoltaic cell, i is the index of the historical time period, and i ∈ [1, Q], Q is the total number of historical time periods.

[0017] Further, feature extraction is performed on the image data to obtain a crack severity index of the surface of the photovoltaic panel, and the specific process is as follows:

[0018] The photovoltaic panel image is converted into a gray-scale image. For an RGB image, conversion is performed through the following formula:

[0019] Gray = 0.299R + 0.587G + 0.114B

[0020] where Gray is the gray value of a pixel, R, G, and B are the pixel values of the red, green, and blue channels, respectively;

[0021] The Sobel operator is used to calculate the gradients in the horizontal and vertical directions:

[0022]

[0023] where Hx is the gradient in the horizontal direction, Hy is the gradient in the vertical direction, I is the photovoltaic panel image, and * represents the convolution operation. x y

[0024] The gradient magnitude and direction are as follows:

[0025]

[0026] where H is the gradient magnitude, and θ is the gradient direction.

[0027] The gradient magnitude H is compared with two thresholds to identify the edge pixels in the image, and the specific process is as follows:

[0028] When H > T high , the pixel is marked as a strong edge pixel.

[0029] When T low ≤ H ≤ T high , the pixel is marked as a weak edge pixel.

[0030] When H < T low , the pixel is marked as a non-edge pixel.

[0031] where T high is the high threshold, and T low is the low threshold.

[0032] ​​From the strong edge pixels, the weak edge pixels connected with the strong edge pixels are marked as edge pixels by using connectivity analysis, so that a complete and continuous edge image is obtained, and the crack profile is extracted based on the edge image.

[0033] Further, for each crack profile detected, the crack severity index of the photovoltaic panel surface is calculated according to the following formula:

[0034]

[0035] wherein, LWZS i is the crack severity index of the photovoltaic panel surface in the i-th historical time period, and the crack severity index is used to evaluate the severity of the crack on the surface of the photovoltaic panel in the time period;

[0036] wherein, A max,i is the maximum crack area of the photovoltaic panel surface in the i-th historical time period, A total is the area of the region in the image corresponding to the crack with the maximum area;

[0037] The temperature deviation index of the photovoltaic panel surface is calculated according to the following formula:

[0038]

[0039] wherein, TDI i is the temperature deviation index of the photovoltaic panel surface in the i-th historical time period, T i is the average temperature of the photovoltaic panel surface in the i-th historical time period, and T0 is the ideal temperature for the normal operation of the photovoltaic panel;

[0040] Further, the open-circuit voltage ratio and the short-circuit current ratio of the photovoltaic cell are processed to generate a cell comprehensive index according to the following formula:

[0041] BATZS i = ω1OVR i + ω2SCR i

[0042] wherein, BATZS i is the cell comprehensive index in the i-th historical time period, and the cell comprehensive index is used to evaluate the health condition of the photovoltaic cell based on the open-circuit voltage ratio and the short-circuit current ratio.

[0043] wherein, ω1 is the weight coefficient of the open-circuit voltage ratio, ω2 is the weight coefficient of the short-circuit current ratio, and ω1 + ω2 = 1, and ω1 = ω2 = 0.5.

[0044] Further, the crack severity index and the temperature deviation index of the surface of the photovoltaic panel are data processed to generate a photovoltaic panel comprehensive index, and the formula is as follows:

[0045] PVZS i =γ1LWZS i +γ2TDI i

[0046] Wherein, PVZS i is the photovoltaic panel comprehensive index of the i-th historical time period, the photovoltaic panel comprehensive index is evaluated from the crack severity index and the temperature deviation index, and the thermal stability of the photovoltaic panel is comprehensively evaluated;

[0047] In the formula, γ1 is the weight coefficient of the crack severity index, γ2 is the weight coefficient of the temperature deviation index, on the basis of γ1+γ2=1, 0<γ2<γ1<1 is satisfied.

[0048] To achieve the above object, the application further provides the following technical scheme:

[0049] An IV, CV fusion diagnosis system of a photovoltaic power station, the system is used for executing any one of the above-mentioned photovoltaic power station IV, CV fusion diagnosis method, and comprises:

[0050] A data acquisition module is used for acquiring the operation data and the operation state of the photovoltaic power station to be diagnosed in a historical time period, the operation data comprises IV data, image data and infrared thermal imaging data, and the operation state comprises one or more state combinations of normal, hot spot, hidden crack, power attenuation and poor welding point contact;

[0051] A feature extraction module is used for performing feature extraction on the IV data, the image data and the infrared thermal imaging data to obtain the feature parameters of the photovoltaic power station to be diagnosed in the historical time period, including the open-circuit voltage ratio of the photovoltaic cell, the short-circuit current ratio, the crack severity index of the surface of the photovoltaic panel and the temperature deviation index;

[0052] A data calculation module is used for data processing the open-circuit voltage ratio and the short-circuit current ratio of the photovoltaic cell to generate a cell comprehensive index for evaluating the health condition of the photovoltaic cell, and data processing the crack severity index and the temperature deviation index of the surface of the photovoltaic panel to generate a photovoltaic panel comprehensive index for evaluating the thermal stability of the photovoltaic panel;

[0053] An identification model construction module is used for constructing an operation state identification model, taking the cell comprehensive index and the photovoltaic panel comprehensive index of each historical time period of the photovoltaic power station to be diagnosed as input and the operation state as label, and training the operation state identification model;

[0054] A test set construction module is configured to acquire the battery comprehensive index and the photovoltaic panel comprehensive index of the photovoltaic power station to be diagnosed at a current time period, input the trained operation state recognition model, and acquire the operation state of the photovoltaic power station to be diagnosed at the current time period.

[0055] Compared with the prior art, the present application has the following advantages:

[0056] The present application can comprehensively reflect the operation state of the photovoltaic power station from multiple angles. The IV data reflect the electrical performance of the photovoltaic cell, the image data identify the physical damage such as cracks on the surface of the photovoltaic panel, and the infrared thermal imaging data can monitor the thermal related problems. The multiple data complement each other, providing more abundant and reliable basis for accurately judging the operation state of the photovoltaic power station.

[0057] The operation state recognition model that has been fully trained can better cope with complex and variable fault conditions. In actual application, the photovoltaic power station may simultaneously have multiple faults, or the fault manifestation is relatively complex. The model can more accurately judge the current fault type by learning various fault modes in the historical data, avoiding the misjudgment that may be caused by simple matching and fusion results.

[0058] By using the battery comprehensive index and the photovoltaic panel comprehensive index, and combining the trained operation state recognition model, the operation state of the photovoltaic power station to be diagnosed at the current time period can be more accurately acquired, avoiding the inaccurate diagnosis problem that may be caused by simply fusing the two diagnosis results in the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 The figure is a schematic diagram of the overall method of the present application.

[0060] Figure 2 The figure is a block diagram of the module composition of the present application. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application is further described in detail below in combination with specific embodiments.

[0062] It should be noted that the technical terms or scientific terms used in the present application should be the general meanings understood by those skilled in the art unless otherwise defined. The "first", "second" and similar words used in the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Include" or "contain" and similar words mean that the elements or objects before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connected" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to represent relative positional relationships, which may change accordingly when the absolute position of the described object changes.

[0063] Embodiment 1:

[0064] Please refer to Figure 1 The present application provides a technical solution:

[0065] A photovoltaic power station IV, CV fusion diagnosis method, the specific steps include:

[0066] S1. Obtain the running data and running state of the photovoltaic power station to be diagnosed in the historical time period, the running data including IV data, image data and infrared thermal imaging data, the running state including one or more state combinations of normal, hot spot, hidden crack, power attenuation and poor contact of welding point;

[0067] S2. Feature extraction is performed on the IV data, image data and infrared thermal imaging data to obtain the feature parameters of the photovoltaic power station to be diagnosed in the historical time period, including the open circuit voltage ratio, short circuit current ratio of the photovoltaic cell, and the crack severity index and temperature deviation index of the photovoltaic panel surface;

[0068] Among them, the IV data corresponds to the open circuit voltage ratio and the short circuit current ratio, the image data corresponds to the crack severity index of the photovoltaic panel surface, and the infrared thermal imaging data corresponds to the temperature deviation index.

[0069] Among them, the crack severity index of the photovoltaic panel surface refers to the ratio of the area of the largest crack to the area of the photovoltaic panel in the image, and the temperature deviation index refers to the deviation between the highest temperature of the photovoltaic panel surface and the ideal temperature under normal working condition.

[0070] On the basis of the above embodiment, the length of each time period is set to 1 day, the time interval between adjacent time periods is equal, the open circuit voltage ratio and the short circuit current ratio of the photovoltaic cell are calculated, and the formula is as follows:

[0071]

[0072] wherein OVR i is the open-circuit voltage ratio of the photovoltaic cell in the i-th historical time period, SCR i is the short-circuit current ratio of the photovoltaic cell in the i-th historical time period, is the actual open-circuit voltage of the photovoltaic cell in the i-th historical time period, V oc0 is the standard open-circuit voltage of the photovoltaic cell, is the actual short-circuit current of the photovoltaic cell in the i-th historical time period, V sc0 is the standard short-circuit current of the photovoltaic cell, i is the index of the historical time period, and i∈[1,Q], Q is the total number of historical time periods;

[0073] wherein the actual open-circuit voltage the actual short-circuit current is the average value obtained by multiple measurements in the i-th historical time period;

[0074] Feature extraction is performed on the image data to obtain a crack severity index of the surface of the photovoltaic panel, and the specific process is as follows:

[0075] The photovoltaic panel image is converted into a gray-scale image. For an RGB image, conversion is performed by the following formula:

[0076] Gray = 0.299R + 0.587G + 0.114B

[0077] wherein Gray is the gray-scale value of a pixel, R, G, and B are the pixel values of the red, green, and blue channels, respectively;

[0078] The Sobel operator is used to calculate the gradients in the horizontal and vertical directions:

[0079]

[0080] wherein H x is the gradient in the horizontal direction, H y is the gradient in the vertical direction, I is the photovoltaic panel image, and * represents a convolution operation,

[0081] The gradient magnitude and direction are respectively:

[0082]

[0083]

[0084] wherein H is the gradient magnitude, and θ is the gradient direction;

[0085] The gradient magnitude H is compared with two thresholds to identify the edge pixels in the image, and the specific process is as follows:

[0086] When H > T highpixels are marked as strong edge pixels;

[0087] when T low ≤ H ≤ T high pixels are marked as weak edge pixels;

[0088] when H < T low pixels are marked as non-edge pixels;

[0089] wherein T high is a high threshold value, and T low is a low threshold value;

[0090] Starting from the strong edge pixels, the weak edge pixels connected therewith are also marked as edge pixels by using connectivity analysis to obtain a complete and continuous edge image, and the profile of the crack is extracted based on the edge image, and a closed curve composed of continuous edge pixels in the image is found by using a profile extraction algorithm, which is the profile of the crack in the image.

[0091] wherein the connectivity analysis is a prior art and will not be described in detail here;

[0092] wherein the specific steps of the profile extraction algorithm for extracting the profile of the defect are as follows:

[0093] The input edge image is binarized to ensure that only foreground (edge pixels) and background exist in the image;

[0094] Starting from the boundary of the image, the first edge pixel that has not been visited is found as a starting point;

[0095] The adjacent edge pixels are tracked in a predetermined direction (e.g., clockwise), and the coordinates of the pixels passed are recorded;

[0096] The tracking is continued until the starting point is returned, forming a closed profile;

[0097] The above steps are repeated until all edge pixels are visited, thereby extracting the profile of the crack;

[0098] For each crack profile detected, the crack severity index of the surface of the photovoltaic panel is calculated according to the following formula:

[0099]

[0100] wherein LWZS i is the crack severity index of the surface of the photovoltaic panel in the i-th historical time period, the crack severity index is used to evaluate the severity of the crack on the surface of the photovoltaic panel in the time period, and the greater the crack severity index, the greater the proportion of the crack in the photovoltaic panel, i.e., the more serious the crack;

[0101] wherein A max,iA is the maximum crack area of the photovoltaic panel surface in the i-th historical time period, wherein a plurality of images are continuously collected in each historical time period max,i A is the maximum crack area corresponding to the image collected last in each historical time period total A is the area of the region of the photovoltaic panel in the image corresponding to the crack with the largest area;

[0102] A is the maximum crack area obtained by filling the region surrounded by the edge according to the contour of the crack, and then calculating the number of pixels of the filled region max,i A is the area of the region of the photovoltaic panel in the image corresponding to the crack with the largest area obtained by the contour extraction and area calculation algorithm total ;

[0103] The temperature deviation index of the surface of the photovoltaic panel is calculated according to the following formula:

[0104]

[0105] TDI i is the temperature deviation index of the surface of the photovoltaic panel in the i-th historical time period, T i is the average temperature of the surface of the photovoltaic panel in the i-th historical time period, and T0 is the ideal temperature for normal operation of the photovoltaic panel;

[0106] A plurality of infrared thermal imaging data are continuously collected in each historical time period, and the mean value is processed to obtain the temperature T i of the surface of the photovoltaic panel.

[0107] On the basis of the above embodiment, the actual open-circuit voltage, the actual short-circuit current, the standard open-circuit voltage, the standard short-circuit current, the temperature of the surface of the photovoltaic panel, and the ideal temperature for normal operation are all mean value processed data.

[0108] On the basis of the above embodiment, after calculating the open-circuit voltage ratio, the short-circuit current ratio, the crack severity index of the surface of the photovoltaic panel, and the temperature deviation index, the data is respectively subjected to maximum-minimum normalization processing, and then the normalized data is used for subsequent analysis and processing, so that various data are analyzed and processed under the same dimension in the subsequent analysis and processing process, and the problem that some data are ignored due to different dimensions is avoided.

[0109] S3. The open-circuit voltage ratio and the short-circuit current ratio of the photovoltaic cell are subjected to data processing to generate a cell comprehensive index for evaluating the health status of the photovoltaic cell, and the crack severity index and the temperature deviation index of the surface of the photovoltaic panel are subjected to data processing to generate a photovoltaic panel comprehensive index for evaluating the thermal stability of the photovoltaic panel;

[0110] On the basis of the above embodiments, the open-circuit voltage ratio and the short-circuit current ratio of the photovoltaic cell are processed to generate a battery comprehensive index, and the formula is as follows:

[0111] BATZS i = ω1OVR i + ω2SCR i

[0112] wherein, BATZS i is the battery comprehensive index of the i-th historical time period, the battery comprehensive index is evaluated from the open-circuit voltage ratio and the short-circuit current ratio, and the battery comprehensive index is greater, the photovoltaic cell is healthier;

[0113] On this basis, it should be noted that: the open-circuit voltage ratio OVR i increases, which means that the battery can output a voltage closer to the ideal value under open-circuit conditions, greatly reducing the voltage drop caused by internal resistance loss and the like, thereby increasing the battery health degree, thereby increasing the battery comprehensive index BATZS i ; the short-circuit current ratio SCR i increases, reflecting the improvement of the photogenerated carrier collection and transport efficiency of the photovoltaic cell, thereby increasing the battery health degree, effectively collecting the photogenerated electron-hole pairs, and in the short-circuit state, the internal charge transport path is unobstructed, and the carrier recombination or trapping by traps is less likely to occur, thereby increasing the battery comprehensive index BATZS i . Therefore, the battery comprehensive index BATZS i is positively correlated with the open-circuit voltage ratio OVR i and the short-circuit current ratio SCR i . Therefore, the above weighted summation formula is used to represent the functional relationship between the battery comprehensive index BATZS i and the open-circuit voltage ratio OVR i and the short-circuit current ratio SCR i .

[0114] In the formula, ω1 is the weight coefficient of the open-circuit voltage ratio, and ω2 is the weight coefficient of the short-circuit current ratio;

[0115] The open-circuit voltage ratio reflects the voltage characteristics of the battery under open-circuit conditions, and the short-circuit current ratio reflects the current characteristics of the battery under short-circuit conditions, both of which are crucial to the performance of the photovoltaic cell. By setting ω1 equal to ω2, the voltage and current performance of the photovoltaic cell can be considered comprehensively and evenly.

[0116] Therefore, on the basis of ω1+ω2=1, let ω1=ω2=0.5.

[0117] On the basis of the above embodiments, the crack severity index and the temperature deviation index of the surface of the photovoltaic panel are processed to generate a comprehensive index of the photovoltaic panel, and the formula is as follows:

[0118] PVZS i =γ1LWZS i +γ2TDI i

[0119] Wherein, PVZS i is the comprehensive index of the photovoltaic panel in the i-th historical time period, the comprehensive index of the photovoltaic panel is evaluated from the crack severity index and the temperature deviation index, and the thermal stability of the photovoltaic panel is worse when the comprehensive index of the photovoltaic panel is larger;

[0120] On this basis, it should be noted that: the crack severity index LWZS i increases, which means that the crack on the surface of the photovoltaic panel is increasingly serious, and the existence and expansion of the crack will damage the structural integrity of the photovoltaic panel, affect the internal heat conduction path, and when the crack severity index LWZS i increases, the thermal resistance of the crack area will increase, and the heat conduction in the photovoltaic panel becomes more uneven, which leads to local temperature rise, so that the thermal stability of the photovoltaic panel becomes worse, thereby increasing the comprehensive index PVZS i of the photovoltaic panel; the temperature deviation index TDI i increases, which means that the gap between the actual average temperature on the surface of the photovoltaic panel and the ideal average temperature under normal working condition is increasing, and no matter the actual temperature is higher or lower than the ideal temperature, the larger temperature deviation will lead to the change of the performance of the photovoltaic panel material, and has a negative impact on the thermal stability of the photovoltaic panel, thereby increasing the comprehensive index PVZS i of the photovoltaic panel. Therefore, the comprehensive index PVZS i of the photovoltaic panel is positively correlated with the crack severity index LWZS i and the temperature deviation index TDI i . Therefore, the above weighted sum formula is used to represent the functional relationship between the comprehensive index PVZS i of the photovoltaic panel and the crack severity index LWZS i and the temperature deviation index TDI i .

[0121] In the formula, γ1 is the weight coefficient of the crack severity index, and γ2 is the weight coefficient of the temperature deviation index;

[0122] Cracks on the surface of the photovoltaic panel will directly destroy the structural integrity and heat conduction path inside it. The presence of cracks may cause heat to concentrate in certain areas, forming hot spots, which in turn triggers a series of thermal problems, seriously affecting the thermal stability of the photovoltaic panel. While temperature deviation also affects thermal stability, it usually has a long-term cumulative effect, which is relatively less direct and significant than the impact of cracks. Therefore, in order to more accurately reflect the thermal stability of the photovoltaic panel, a higher weight is given to the crack severity index, i.e. γ1> γ2.

[0123] Therefore, on the basis of γ1+ γ2=1, let 0< γ2< γ1< 1.

[0124] As an embodiment, the value range of γ1 is 0.5-1, and the value range of γ2 is 0-0.5. The specific value is set by the technical personnel according to the actual situation, which is not limited here.

[0125] S4. Construct a running state recognition model, taking the battery comprehensive index and the photovoltaic panel comprehensive index of each historical time period of the photovoltaic power station as input and the running state as label, train the running state recognition model;

[0126] On the basis of the above embodiment, the running state recognition model is constituted by a deep learning network based on a multilayer perceptron, and the deep neural network of the multilayer perceptron includes an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer. The first hidden layer, the second hidden layer and the third hidden layer each have at least two neurons, and each uses ReLU as an activation function.

[0127] In the running state recognition model, the input features of the deep learning network of the multilayer perceptron include: the battery comprehensive index and the photovoltaic panel comprehensive index, 2 features.

[0128] The structure of the deep learning network of the multilayer perceptron is as follows:

[0129] Input layer: receiving input of 2 features;

[0130] First hidden layer: having 128 neurons, using ReLU as activation function;

[0131] Second hidden layer: having 64 neurons, also using ReLU activation function;

[0132] Third hidden layer: having 32 neurons, using ReLU activation function;

[0133] Output layer: having 1 neuron, outputting running state.

[0134] The process of training the running state recognition model is as follows:

[0135] The battery comprehensive index and the photovoltaic panel comprehensive index of each historical time period of the photovoltaic power station to be diagnosed are taken as inputs, and the running state is taken as a label to train, and the mean square error is taken as a loss function, and when the mean square error is in the range of [0, 0.01], the training of the running state recognition model is completed.

[0136] S5. The battery comprehensive index and the photovoltaic panel comprehensive index of the photovoltaic power station to be diagnosed in the current time period are obtained, the trained running state recognition model is input, and the running state of the photovoltaic power station to be diagnosed in the current time period is obtained.

[0137] Please refer to Figure 2 The application also provides a technical solution:

[0138] A photovoltaic power station IV, CV fusion diagnosis system, the system is used for executing any one of the above-mentioned photovoltaic power station IV, CV fusion diagnosis method, comprising:

[0139] A data acquisition module is configured to acquire running data and a running state of a photovoltaic power station to be diagnosed in a historical time period, wherein the running data comprises IV data, image data and infrared thermal imaging data, and the running state comprises one or more state combinations of normal, hot spot, hidden crack, power attenuation and poor welding point contact.

[0140] A feature extraction module is configured to extract features from the IV data, the image data and the infrared thermal imaging data to obtain feature parameters of the photovoltaic power station to be diagnosed in the historical time period, including an open-circuit voltage ratio of a photovoltaic cell, a short-circuit current ratio, a crack severity index of a photovoltaic panel surface and a temperature deviation index.

[0141] A data calculation module is configured to process the open-circuit voltage ratio and the short-circuit current ratio of the photovoltaic cell to generate a battery comprehensive index for evaluating a health condition of the photovoltaic cell, and process the crack severity index and the temperature deviation index of the photovoltaic panel surface to generate a photovoltaic panel comprehensive index for evaluating thermal stability of the photovoltaic panel.

[0142] An identification model construction module is configured to construct a running state recognition model, take the battery comprehensive index and the photovoltaic panel comprehensive index of each historical time period of the photovoltaic power station to be diagnosed as inputs, and take the running state as a label to train the running state recognition model.

[0143] A test set construction module is configured to obtain the battery comprehensive index and the photovoltaic panel comprehensive index of the photovoltaic power station to be diagnosed in a current time period, input the trained running state recognition model, and obtain the running state of the photovoltaic power station to be diagnosed in the current time period.

[0144] The above formulas are all dimensionless values calculated, the formula is obtained by collecting a large amount of data to simulate the most recent real situation, and the preset parameters in the formula are set by a person skilled in the art according to the actual situation.

[0145] The above embodiments can be implemented wholly or partially by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented wholly or partially in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of the examples described in connection with the embodiments disclosed herein can be realized by electronic hardware, or a combination of computer software and electronic hardware. Whether the functions are executed in hardware or software methods depends on the specific application and design constraints of the technical solutions.

[0146] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, which can be located in one place or distributed on multiple network units. Part or all of the units can be selected to achieve the purpose of the embodiments according to actual needs.

[0147] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. A photovoltaic power station IV, CV fusion diagnosis method, characterized in that: The specific steps include: S1. Obtain the operation data and operation state of the photovoltaic power station to be diagnosed in the historical time period, the operation data including IV data, image data and infrared thermal imaging data, the operation state including normal, or one or more state combinations of hot spot, hidden crack, power attenuation and poor contact of welding point; S2. Feature extraction is performed on the IV data, image data and infrared thermal imaging data to obtain the feature parameters of the photovoltaic power station to be diagnosed in the historical time period, including the open circuit voltage ratio and short circuit current ratio of the photovoltaic cell, and the crack severity index and temperature deviation index of the photovoltaic panel surface; S3. The open circuit voltage ratio and short circuit current ratio of the photovoltaic cell are processed to generate a cell comprehensive index for evaluating the health status of the photovoltaic cell, and the crack severity index and temperature deviation index of the photovoltaic panel surface are processed to generate a photovoltaic panel comprehensive index for evaluating the thermal stability of the photovoltaic panel; S4. An operation state recognition model is constructed, taking the cell comprehensive index and the photovoltaic panel comprehensive index of each historical time period of the photovoltaic power station to be diagnosed as input and the operation state as label, and the operation state recognition model is trained; S5. The cell comprehensive index and the photovoltaic panel comprehensive index of the photovoltaic power station to be diagnosed in the current time period are obtained, input into the trained operation state recognition model, and the operation state of the photovoltaic power station to be diagnosed in the current time period is obtained; The image data is subjected to feature extraction, and the specific process is as follows: The photovoltaic panel image is converted into a gray image, and for an RGB image, the conversion is performed through the following formula: wherein, is a gray value of a pixel, , , is a pixel value of a red channel, a green channel, a blue channel. The Sobel operator is used to calculate the gradient in the horizontal and vertical directions: wherein is a gradient in horizontal direction, is a gradient in vertical direction, is a photovoltaic panel image, denotes a convolution operation, The gradient amplitude and direction are respectively: wherein is the gradient magnitude, is the gradient direction; The gradient magnitude and two threshold values, which identifies the edge pixels in the image, the specific process is as follows: When , the pixel is marked as a strong edge pixel; When , the pixel is marked as a weak edge pixel; When , the pixel is marked as a non-edge pixel; wherein is a high threshold value, is a low threshold value; Starting from the strong edge pixels, the weak edge pixels connected thereto are also marked as edge pixels through connectivity analysis to obtain a complete and continuous edge image, and the crack contour is extracted based on the edge image. The closed curve composed of continuous edge pixels in the image is found through the contour extraction algorithm, which is the crack contour of the image.

2. The photovoltaic power station IV, CV fusion diagnosis method according to claim 1, characterized in that: The length of each time period is set to 1 day, the time interval between adjacent time periods is equal, the open circuit voltage ratio and short circuit current ratio of the photovoltaic cell are calculated, and the formula is as follows: wherein is the open circuit voltage ratio of the photovoltaic cell for the th historical time period, is the short circuit current ratio of the photovoltaic cell for the th historical time period, is the actual open circuit voltage of the photovoltaic cell for the th historical time period, is the standard open circuit voltage of the photovoltaic cell, is the actual short circuit current of the photovoltaic cell for the th historical time period, is the standard short circuit current of the photovoltaic cell, is an index of the historical time period, and , is the total number of historical time periods.

3. The photovoltaic power station IV, CV fusion diagnosis method according to claim 1, characterized in that: For each crack contour detected, the crack severity index of the photovoltaic panel surface is calculated according to the following formula: wherein, is the number of the first historical time period, is the crack severity index of the photovoltaic panel surface in the first historical time period, the crack severity index being used to evaluate the severity of the cracks on the photovoltaic panel surface in the first historical time period; In the formula, is the maximum crack area of the photovoltaic panel surface in the first historical time period, is the maximum crack area of the photovoltaic panel surface in the first historical time period, is the area of the region in the image corresponding to the crack with the largest area. The temperature deviation index of the photovoltaic panel surface is calculated according to the following formula: wherein, is the temperature deviation index of the photovoltaic panel surface for the first historical time period, is the temperature deviation index of the photovoltaic panel surface for the first historical time period, is the average temperature of the photovoltaic panel surface for the first historical time period, is the average temperature of the photovoltaic panel surface for the first historical time period, is the ideal temperature for the normal operation of the photovoltaic panel.

4. The photovoltaic power station IV, CV fusion diagnosis method according to claim 2, characterized in that: The open circuit voltage ratio and short circuit current ratio of the photovoltaic cell are processed to generate a cell comprehensive index, and the formula is as follows: in, For the first The battery comprehensive index for a historical period, which comprehensively assesses the health status of photovoltaic cells based on two indicators: open-circuit voltage ratio and short-circuit current ratio. wherein is a weight coefficient for the open-circuit voltage ratio, is a weight coefficient for the short-circuit current ratio, and is based on the condition that .

5. The photovoltaic power station IV, CV fusion diagnosis method according to claim 3, characterized in that: The crack severity index and temperature deviation index of the photovoltaic panel surface are processed to generate a photovoltaic panel comprehensive index, and the formula is as follows: in, For the first The comprehensive index of photovoltaic panels over a historical period is used to comprehensively assess the thermal stability of photovoltaic panels based on two indicators: crack severity index and temperature deviation index. In the formula, is a weight coefficient of the crack severity index, is a weight coefficient of the temperature deviation index, and is a weight coefficient of the crack severity index, is a weight coefficient of the crack severity index.

6. An IV, CV fusion diagnosis system for photovoltaic power station, the system is used to execute the method of any one of claims 1-5, characterized in that: It includes: A data acquisition module is used to obtain the operation data and operation state of the photovoltaic power station to be diagnosed in the historical time period, the operation data including IV data, image data and infrared thermal imaging data, the operation state including normal, or one or more state combinations of hot spot, hidden crack, power attenuation and poor contact of welding point; The feature extraction module is configured to perform feature extraction on the IV data, the image data, and the infrared thermal imaging data to obtain feature parameters of the photovoltaic power station to be diagnosed in historical time periods, including an open-circuit voltage ratio of the photovoltaic cell, a short-circuit current ratio, a crack severity index of a surface of the photovoltaic panel, and a temperature deviation index. The data calculation module is configured to perform data processing on the open-circuit voltage ratio and the short-circuit current ratio of the photovoltaic cell to generate a cell comprehensive index for evaluating a health condition of the photovoltaic cell, and perform data processing on the crack severity index and the temperature deviation index of the surface of the photovoltaic panel to generate a photovoltaic panel comprehensive index for evaluating thermal stability of the photovoltaic panel. The recognition model construction module is configured to construct a running state recognition model, take the cell comprehensive index and the photovoltaic panel comprehensive index of each historical time period of the photovoltaic power station to be diagnosed as input, and take a running state as a label, and train the running state recognition model. The test set construction module is configured to obtain the cell comprehensive index and the photovoltaic panel comprehensive index of the photovoltaic power station to be diagnosed in a current time period, input the trained running state recognition model, and obtain a running state of the photovoltaic power station to be diagnosed in the current time period.

Images