Photovoltaic array fault diagnosis method, device, equipment, medium and program product

Abstract

The invention discloses a photovoltaic array fault diagnosis method, device and equipment, a medium and a program product, relates to the technical field of photovoltaic power generation, and discloses a photovoltaic array fault diagnosis method comprising the following steps: obtaining characteristic parameters of a photovoltaic array; wherein the characteristic parameters at least comprise a maximum power point voltage, a maximum power point current and a fill factor; based on the maximum power point voltage, the maximum power point current and the fill factor, performing fault diagnosis on the characteristic parameters of the photovoltaic array by using a pre-trained integrated support vector machine model to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained by combining a support vector machine with integrated learning training. According to the technical scheme, the integrated support vector machine model has the advantages of the support vector machine in the aspect of classification and the characteristics of integrated learning in the aspect of integrating the advantages of a plurality of weak classifiers at the same time, so that the integrated support vector machine model has excellent diagnosis accuracy in the aspect of fault diagnosis of the photovoltaic array.

Description

Technical Field

[0001] The present application relates to the field of photovoltaic power generation technology, and in particular to photovoltaic array fault diagnosis methods, devices, equipment, media and program products. Background Art

[0002] At present, photovoltaic power stations are increasingly used on deep-sea platforms. On the "Hengyi No. 1" deep-sea aquaculture platform designed by Zhanjiang Bay Laboratory, the photovoltaic power station system has met the daily production and living electricity needs of the platform, which is not only environmentally friendly, but also saves a lot of fuel consumption. In the future, photovoltaic power stations will play an increasingly important role on deep-sea platforms.

[0003] On deep-sea platforms, photovoltaic modules operate in outdoor offshore environments for a long time. Their power generation performance is affected by meteorological factors, especially solar irradiance and ambient temperature and humidity. There will be faults such as model aging and hot spot effects, which will lead to reduced power generation efficiency of the system.

[0004] At present, most photovoltaic system fault diagnosis methods use simulation software. When creating these simulation models, it is still difficult to ensure high accuracy and scalability of the models. In addition, the established models are also difficult to be consistent with actual working conditions to a large extent. Once the model is more complex, the difficulty of modeling will be greatly increased.

[0005] Therefore, a new photovoltaic array fault diagnosis method is urgently needed to solve the problem that the existing photovoltaic array fault diagnosis methods mostly use simulation software to establish simulation models, which have high modeling difficulty and insufficient accuracy and are difficult to meet fault diagnosis needs. Summary of the invention

[0006] The main purpose of the present application is to provide a photovoltaic array fault diagnosis method, device, equipment, medium and program product, aiming to solve the technical problem that the existing photovoltaic array fault diagnosis methods mostly use simulation software to establish simulation models, which have high modeling difficulty and insufficient accuracy and are difficult to meet the fault diagnosis needs.

[0007] To achieve the above objectives, the present application proposes a photovoltaic array fault diagnosis method, the method comprising:

[0008] Acquiring characteristic parameters of a photovoltaic array; wherein the characteristic parameters include at least a maximum power point voltage, a maximum power point current and a fill factor;

[0009] Based on the maximum power point voltage, the maximum power point current and the fill factor, a pre-trained integrated support vector machine model is used to perform fault diagnosis on characteristic parameters of the photovoltaic array to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

[0010] In one embodiment, after the step of obtaining characteristic parameters of the photovoltaic array, the following steps are included:

[0011] Normalizing the maximum power point voltage and the maximum power point current to obtain a normalized maximum power point voltage and a normalized maximum power point current;

[0012] The normalized maximum power point voltage and the normalized maximum power point current are added to the characteristic parameters.

[0013] In one embodiment, before the step of performing fault diagnosis on characteristic parameters of the photovoltaic array using a pre-trained integrated support vector machine model based on the maximum power point voltage, the maximum power point current and the fill factor to obtain a fault diagnosis result of the photovoltaic array, the method comprises:

[0014] Based on the characteristic parameters, preparing sample data sets of multiple categories, and dividing the sample data sets into a training subset and a test subset;

[0015] Initializing the weights of the training samples in the training subset to obtain initialized sample weights;

[0016] Based on the training subset and the initialization sample weight, multiple weak classifiers are trained respectively to obtain multiple target weak classifiers; wherein the weak classifier is a model based on a support vector machine;

[0017] A plurality of the target weak classifiers are combined into the integrated support vector machine model.

[0018] In one embodiment, the step of training a plurality of weak classifiers based on the training subset and the initialization sample weights to obtain a plurality of target weak classifiers comprises:

[0019] Based on the sample distribution, n training samples are selected from the training subset, and the weights of the training samples are normalized to obtain normalized sample weights;

[0020] A training step, based on the training samples and the normalized sample weights, training a plurality of the weak classifiers according to a preset number of training rounds; during the training process of the weak classifiers, calculating the classification error rate of the weak classifiers, and calculating and updating the weight of each of the weak classifiers and the weight of the training samples;

[0021] If the classification error rate is less than or equal to a first preset threshold, the iteration is stopped to obtain the target weak classifier;

[0022] If the classification error rate is greater than or equal to the second preset threshold, stop the iteration, adjust the normalized sample weight, and return to execute the training step until the classification error rate of the weak classifier finally trained is less than or equal to the first preset threshold;

[0023] If the classification error rate is any other case, return to execute the training step until the classification error rate of the weak classifier finally obtained by training is less than or equal to the first preset threshold.

[0024] In one embodiment, the parameters of the support vector machine in the weak classifier include at least a kernel function, and the kernel function is set to a Gaussian kernel function.

[0025] In one embodiment, the classification result output by the integrated support vector machine model is:

[0026]

[0027] Wherein, H(x) represents the classification result output by the integrated support vector machine model; sgn(*) represents the sign function; α i represents the weight of the i-th target weak classifier; f i (x i ) represents the i-th target weak classifier in the i-th sample data x i The output value on .

[0028] In addition, to achieve the above-mentioned purpose, the present application also proposes a photovoltaic array fault diagnosis device, the photovoltaic array fault diagnosis device comprising:

[0029] A parameter acquisition module, used to acquire characteristic parameters of the photovoltaic array; wherein the characteristic parameters at least include maximum power point voltage, maximum power point current and fill factor;

[0030] A fault diagnosis module is used to perform fault diagnosis on characteristic parameters of the photovoltaic array based on the maximum power point voltage, the maximum power point current and the fill factor using a pre-trained integrated support vector machine model to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

[0031] In addition, to achieve the above-mentioned purpose, the present application also proposes a photovoltaic array fault diagnosis device, the device comprising: a data acquisition module, a data processing and storage module, a fault diagnosis module and a communication module;

[0032] The data acquisition module is used to collect power generation information of the photovoltaic array and transmit the power generation information to the data processing and storage module;

[0033] The data processing and storage module is used to pre-process the power generation information to obtain the characteristic parameters in the photovoltaic array fault diagnosis method as described above;

[0034] The fault diagnosis module is used to store and run a computer program; wherein the computer program is configured to implement the steps of the photovoltaic array fault diagnosis method as described above;

[0035] The communication module is used to transmit the power generation information and the fault diagnosis result to the host computer main control module.

[0036] In addition, to achieve the above objectives, the present application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the photovoltaic array fault diagnosis method described above are implemented.

[0037] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the photovoltaic array fault diagnosis method described above are implemented.

[0038] One or more technical solutions proposed in this application have at least the following technical effects:

[0039] The present application provides a photovoltaic array fault diagnosis method. The present application first obtains characteristic parameters of a photovoltaic array; wherein the characteristic parameters include at least a maximum power point voltage, a maximum power point current and a fill factor; based on the maximum power point voltage, the maximum power point current and the fill factor, a pre-trained integrated support vector machine model is used to perform fault diagnosis on the characteristic parameters of the photovoltaic array to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on a support vector machine combined with integrated learning training.

[0040] The technical solution of the present application utilizes an integrated support vector machine model to process the characteristic parameters of a photovoltaic array. Since the integrated support vector machine model is obtained based on support vector machines combined with integrated learning training, it has both the advantages of support vector machines in classification problems and the characteristics of integrated learning technology that can combine the advantages of weak classifiers and eliminate the shortcomings of single weak classifiers such as overfitting, thereby improving the accuracy and stability of the classifier. The integrated support vector machine model obtained by combining support vector machines and integrated learning technology has excellent diagnostic accuracy in fault diagnosis of photovoltaic arrays and improves diagnostic efficiency.

[0041] In addition, compared with the prior art solution that only uses the maximum power point voltage and the maximum power point current as characteristic parameters, the present application also additionally uses the fill factor as a characteristic parameter of the photovoltaic array; since the fill factor can reflect the gap between the actual performance of the photovoltaic array and the theoretical optimal performance, the maximum power point voltage, the maximum power point current and the fill factor are used as characteristic parameters for fault diagnosis, which can comprehensively combine the characteristics of various parameters to more comprehensively diagnose the performance status of the entire photovoltaic array and improve the accuracy of fault type diagnosis; and after using the fill factor as a characteristic parameter, it has a good diagnostic effect on some specific faults that will not significantly change the maximum power point voltage or current but will affect the fill factor. For example, the mismatch between the cells of the photovoltaic array or the problem of partial shadowing may reduce the fill factor, while having little impact on the maximum power point voltage and the maximum power point current. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0043] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0044] Figure 1 A schematic diagram of a flow chart provided for the first embodiment of the photovoltaic array fault diagnosis method of the present application;

[0045] Figure 2 A schematic diagram of a flow chart provided for the second embodiment of the photovoltaic array fault diagnosis method of the present application;

[0046] Figure 3 A schematic diagram of a flow chart provided for the third embodiment of the photovoltaic array fault diagnosis method of the present application;

[0047] Figure 4 A schematic diagram of a brief flow chart of a photovoltaic array fault diagnosis method provided in the third embodiment of the present application;

[0048] Figure 5 This is a schematic diagram of the module structure of the photovoltaic array fault diagnosis device according to an embodiment of the present application;

[0049] Figure 6 Schematic diagram of the device structure of the hardware operating environment involved in the photovoltaic array fault diagnosis method in the embodiment of the present application.

[0050] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0051] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0052] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0053] The main solution of the embodiment of the present application is: obtaining characteristic parameters of a photovoltaic array; wherein the characteristic parameters include at least a maximum power point voltage, a maximum power point current and a fill factor; based on the maximum power point voltage, the maximum power point current and the fill factor, using a pre-trained integrated support vector machine model to perform fault diagnosis on the characteristic parameters of the photovoltaic array to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

[0054] In this embodiment, for the convenience of description, the photovoltaic array fault diagnosis device is used as the execution subject for explanation below.

[0055] At present, photovoltaic power stations are increasingly used on deep-sea platforms. On the "Hengyi No. 1" deep-sea aquaculture platform designed by Zhanjiang Bay Laboratory, the photovoltaic power station system has met the daily production and living electricity needs of the platform, which is not only environmentally friendly, but also saves a lot of fuel consumption. In the future, photovoltaic power stations will play an increasingly important role on deep-sea platforms.

[0056] On deep-sea platforms, photovoltaic modules operate in outdoor offshore environments for a long time. Their power generation performance is affected by meteorological factors, especially solar irradiance and ambient temperature and humidity. There will be faults such as model aging and hot spot effects, which will lead to reduced power generation efficiency of the system.

[0057] At present, most photovoltaic system fault diagnosis methods use simulation software. When creating these simulation models, it is still difficult to ensure high accuracy and scalability of the models. In addition, the established models are also difficult to be consistent with actual working conditions to a large extent. Once the model is more complex, the difficulty of modeling will be greatly increased.

[0058] Therefore, a new photovoltaic array fault diagnosis method is urgently needed to solve the problem that the existing photovoltaic array fault diagnosis methods mostly use simulation software to establish simulation models, which have high modeling difficulty and insufficient accuracy and are difficult to meet fault diagnosis needs.

[0059] Based on this, an embodiment of the present application proposes a solution. The present application first obtains characteristic parameters of a photovoltaic array; wherein the characteristic parameters include at least a maximum power point voltage, a maximum power point current and a fill factor; based on the maximum power point voltage, the maximum power point current and the fill factor, a pre-trained integrated support vector machine model is used to perform fault diagnosis on the characteristic parameters of the photovoltaic array to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

[0060] The technical solution of the present application utilizes an integrated support vector machine model to process the characteristic parameters of a photovoltaic array. Since the integrated support vector machine model is obtained based on support vector machines combined with integrated learning training, it has both the advantages of support vector machines in classification problems and the characteristics of integrated learning technology that can combine the advantages of weak classifiers and eliminate the shortcomings of single weak classifiers such as overfitting, thereby improving the accuracy and stability of the classifier. The integrated support vector machine model obtained by combining support vector machines and integrated learning technology has excellent diagnostic accuracy in fault diagnosis of photovoltaic arrays and improves diagnostic efficiency.

[0061] In addition, compared with the prior art solution that only uses the maximum power point voltage and the maximum power point current as characteristic parameters, the present application also additionally uses the fill factor as a characteristic parameter of the photovoltaic array; since the fill factor can reflect the gap between the actual performance of the photovoltaic array and the theoretical optimal performance, the maximum power point voltage, the maximum power point current and the fill factor are used as characteristic parameters for fault diagnosis, which can comprehensively combine the characteristics of various parameters to more comprehensively diagnose the performance status of the entire photovoltaic array and improve the accuracy of fault type diagnosis; and after using the fill factor as a characteristic parameter, it has a good diagnostic effect on some specific faults that will not significantly change the maximum power point voltage or current but will affect the fill factor. For example, the mismatch between the cells of the photovoltaic array or the problem of partial shadowing may reduce the fill factor, while having little impact on the maximum power point voltage and the maximum power point current.

[0062] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a photovoltaic array fault diagnosis device, etc. The following takes the photovoltaic array fault diagnosis device as an example to illustrate this embodiment and the following embodiments.

[0063] Based on this, the present application embodiment provides a photovoltaic array fault diagnosis method, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the photovoltaic array fault diagnosis method of the present application.

[0064] In this embodiment, the photovoltaic array fault diagnosis method includes steps S10 to S80:

[0065] Step S10, obtaining characteristic parameters of the photovoltaic array; wherein the characteristic parameters at least include maximum power point voltage, maximum power point current and fill factor;

[0066] It should be noted that the maximum power point refers to the maximum power point that the photovoltaic array can output. The maximum power point voltage (V mpp ), maximum power point current (I mpp ) is the voltage and current corresponding to the photovoltaic array when it reaches the maximum power point. The fill factor refers to the maximum power of the photovoltaic array and the open circuit voltage (V oc ), short circuit current (I sc ) times the product of .

[0067] It is understandable that, since the maximum power point voltage, the maximum power point current and the fill factor reflect different characteristics of the photovoltaic array respectively, performing step S10 can facilitate a more comprehensive subsequent diagnosis of the performance status of the entire photovoltaic array, thereby improving the accuracy of diagnosing the fault type of the photovoltaic array.

[0068] Step S80, based on the maximum power point voltage, the maximum power point current and the fill factor, use a pre-trained integrated support vector machine model to perform fault diagnosis on the characteristic parameters of the photovoltaic array to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

[0069] It should be noted that Support Vector Machine (SVM) is a supervised learning model in machine learning technology, mainly used for classification and regression analysis. Its basic principle is to find an optimal classification hyperplane for a linearly separable data set to achieve the division of sample data of different categories. The selection criteria of the optimal hyperplane are to minimize the training error while maximizing the classification interval, which refers to the distance between the nearest positive and negative samples and the hyperplane, and the goal of the support vector machine is to maximize this distance, so as to better distinguish sample data of different categories.

[0070] It should also be noted that ensemble learning refers to a machine learning technology that improves prediction performance by combining the prediction results of multiple basic models (i.e., weak classifiers). Specifically, ensemble learning reduces bias and variance by combining the prediction results of multiple weak classifiers, thereby improving the overall prediction accuracy, and diversified weak classifiers help reduce overfitting. Therefore, even if some weak classifiers are overfitted, other weak classifiers can compensate for this bias, which is suitable for classification tasks such as fault diagnosis of photovoltaic arrays.

[0071] This embodiment provides a photovoltaic array fault diagnosis method, which processes the characteristic parameters of the photovoltaic array by using an integrated support vector machine model. Since the integrated support vector machine model is obtained based on the support vector machine combined with integrated learning training, it has both the advantages of the support vector machine in classification problems and the characteristics of the integrated learning technology that can combine the advantages of multiple weak classifiers and eliminate the shortcomings of a single weak classifier such as overfitting, thereby improving the accuracy and stability of the classifier. The integrated support vector machine model obtained by combining the support vector machine and the integrated learning technology has excellent diagnostic accuracy in the fault diagnosis of the photovoltaic array and improves the diagnostic efficiency.

[0072] In addition, compared with the prior art solution that only uses the maximum power point voltage and the maximum power point current as characteristic parameters, the present application also additionally uses the fill factor as a characteristic parameter of the photovoltaic array; since the fill factor can reflect the gap between the actual performance of the photovoltaic array and the theoretical optimal performance, the maximum power point voltage, the maximum power point current and the fill factor are used as characteristic parameters for fault diagnosis, which can comprehensively combine the characteristics of various parameters to more comprehensively diagnose the performance status of the entire photovoltaic array and improve the accuracy of fault type diagnosis; and after using the fill factor as a characteristic parameter, it has a good diagnostic effect on some specific faults that will not significantly change the maximum power point voltage or current but will affect the fill factor. For example, the mismatch between the cells of the photovoltaic array or the problem of partial shadowing may reduce the fill factor, while having little impact on the maximum power point voltage and the maximum power point current.

[0073] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the first embodiment can refer to the above introduction, and will not be repeated in the following. Figure 2 After step S10, the photovoltaic array fault diagnosis method further includes steps S20 to S30:

[0074] Step S20, normalizing the maximum power point voltage and the maximum power point current to obtain a normalized maximum power point voltage and a normalized maximum power point current;

[0075] It should be noted that normalization refers to a data preprocessing technique that is used to convert data into a specific range or standardize it to a form with a mean of 0 and a variance of 1.

[0076] Specifically, in step S20, the normalization processing method for the maximum power point voltage is as follows:

[0077]

[0078] Among them, V norm Represents the normalized maximum power point voltage; V mpp represents the maximum power point voltage; M represents the number of panels in each photovoltaic array branch; V oc-ref Represents the open circuit voltage of the reference PV array.

[0079] The normalization method for the maximum power point current is as follows:

[0080]

[0081] Among them, I norm Represents the normalized maximum power point current; I mpp Indicates the maximum power point current; N indicates the number of branches of the photovoltaic array; I sc-ref Indicates the reference PV array short-circuit current.

[0082] Correspondingly, the calculation method of the filling factor FF can refer to the following formula:

[0083]

[0084] Among them, I mpp Indicates the maximum power point current; V mpp represents the maximum power point voltage; N represents the number of branches of the photovoltaic array; I sc-ref represents the reference PV array short-circuit current; M represents the number of panels in each PV array branch; V oc-ref Represents the open circuit voltage of the reference PV array.

[0085] In step S20, by normalizing the maximum power point voltage and the maximum power point current, different photovoltaic arrays or panels are converted into values ​​with the same reference standard, thereby helping the integrated support vector machine model to better analyze changes in data distribution and improve the fault diagnosis accuracy of the model.

[0086] Step S30: adding the normalized maximum power point voltage and the normalized maximum power point current to the characteristic parameters.

[0087] At this time, correspondingly, step S80 becomes: based on the maximum power point voltage, the maximum power point current, the normalized maximum power point voltage, the normalized maximum power point current and the fill factor, a pre-trained integrated support vector machine model is used to perform fault diagnosis on the characteristic parameters of the photovoltaic array to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

[0088] In the second embodiment of the present application, by normalizing the maximum power point voltage and the maximum power point current, different photovoltaic arrays or panels are converted into numerical values ​​with the same reference standard, and the normalized maximum power point voltage and the normalized maximum power point current are added to the characteristic parameters, which helps the subsequent integrated support vector machine model to better analyze the characteristic parameters and improve the accuracy of the integrated support vector machine model in fault diagnosis of photovoltaic arrays.

[0089] Based on the first embodiment or the second embodiment of the present application, in the third embodiment of the present application, the same or similar contents as those of the first embodiment or the second embodiment can refer to the above introduction, and will not be repeated later. Figure 3 and Figure 4 Before step S80, the photovoltaic array fault diagnosis method further includes steps S40 to S70:

[0090] Step S40, based on the feature parameters, preparing sample data sets of multiple categories, and dividing the sample data sets into a training subset and a test subset;

[0091] It should be noted that the categories of the sample data set can be divided according to different fault types or according to other requirements. The sample data set can be divided into a training subset and a test subset according to a specific division ratio, which can be 7:3 or 9:1. For large sample data sets, the division ratio can be further increased, such as 99:1, so that the model can learn as many features as possible from the training subset; the specific division ratio can be selected according to the size of the sample data set and the training requirements, and is not limited here.

[0092] In step S40, for n training samples, there are corresponding sample data (x1, y1), ..., (x n ,y n ), where x i ∈X is the sample feature in the sample space, y i ∈Y is the category label in the category space.

[0093] Step S50, initializing the weights of the training samples in the training subset to obtain the initialized sample weights;

[0094] Among them, for the sample data (x i ,y i ) is D(i), and the initial distribution values ​​of all samples are set to 1 / n, where n is the number of training samples. The initialized sample weight is w1(i):

[0095]

[0096] Step S60, based on the training subset and the initialization sample weight, respectively train multiple weak classifiers to obtain multiple target weak classifiers; wherein the weak classifier is a model based on a support vector machine;

[0097] It should be noted that the parameters of the support vector machine in the weak classifier include at least a kernel function. As an optional implementation, the kernel function can be set to a Gaussian kernel function. The role of the kernel function is to map the sample data from the original feature space to a high-dimensional space by using the kernel function when the sample data in the training subset is not linearly separable. The sample data mapped to the high-dimensional space can find the corresponding most classified hyperplane to distinguish sample data of different categories. Kernel functions include linear kernel functions, polynomial kernel functions, Gaussian kernel functions, etc. The Gaussian kernel function is used in this application. The Gaussian kernel function can adapt to various complex decision boundaries. Therefore, it has strong adaptability to tasks that require high precision, such as fault diagnosis of photovoltaic arrays, and helps to diagnose the fault type more accurately.

[0098] Specifically, the construction process of the support vector machine can refer to the following method: For a given training sample data set D = {x i ,y i}, i = 1, 2, 3, ..., m; y i ∈{-1,1}, where x i is the sample data, m is the total number of training samples, y i is the label corresponding to the sample. The sample data in the training sample data set can be separated by an optimal classification hyperplane, which can be expressed as w T x+b=0, where w is the normal vector, w∈R d ; d is the dimension of the sample space, which determines the direction of the hyperplane; b is the displacement term, b∈R, which determines the distance between the hyperplane and the origin. Assuming that the hyperplane can correctly classify the training samples, for {x i ,y i}∈D, if y i = +1, then w T x+b>0; if y i =-1, then w T x+b<0.

[0099] Then the number of training sample points closest to the hyperplane makes the equality of the above equation hold. Then the sum of the distances of the support vectors of two different categories to the hyperplane, that is, the classification interval, is:

[0100] γ=2 / ||w|| 2

[0101] Among them, ||w|| is the L2 norm of w.

[0102] To maximize γ, we need ||w|| 2 Minimum, so the optimal classification problem can be converted into a quadratic programming problem, that is, to find the minimum value of the function:

[0103]

[0104] Make y i ((w·x i )+b)≥1, i=1,...,l

[0105] The minimum value is solved by seeking the dual problem of the above formula. For this purpose, the Lagrange function can be introduced:

[0106]

[0107] Among them, α i is the Lagrange coefficient, α i >0.

[0108] Then find the minimum value of the Lagrange function for w and b. To avoid calculating sample x i and x j The inner product operation in high-dimensional space is calculated using the Gaussian kernel function.

[0109] Based on the above implementation, a support vector machine can be constructed, and the model based on the support vector machine can be trained as a weak classifier.

[0110] In one feasible implementation, step S60 may include steps S61 to S65:

[0111] Step S61, based on sample distribution, select n training samples from the training subset, and normalize the weights of the training samples to obtain normalized sample weights;

[0112] The normalization process of sample weights can be referred to the following formula:

[0113]

[0114] Among them, w t (i) represents the normalized sample weight.

[0115] In step S61, the weights of the training samples are normalized to help the weak classifier reduce computational complexity and improve computational efficiency.

[0116] Step S62, a training step, based on the training samples and the normalized sample weights, training a plurality of the weak classifiers according to a preset number of training rounds; during the training process of the weak classifiers, calculating the classification error rate of the weak classifiers, and calculating and updating the weight of each of the weak classifiers and the weight of the training samples;

[0117] In step S62, the training process calculates the classification error rate of the weak classifier in the following manner:

[0118]

[0119] Among them, ε t represents the classification error rate; w t (i) represents the weight of the training sample; f i (x i ) represents the i-th weak classifier in the i-th sample data x i Output value on y i represents the label of the sample data; I(*) represents the indicator function, I(f i (x i )≠y i ) means when f i (x i )≠y i When f i (x i )=y i , the output value of the indicator function is 0.

[0120] The way to calculate the weights of the weak classifier in the training process can be referred to as follows:

[0121]

[0122] Among them, α i represents the classifier weight, ε t Represents the classification error rate.

[0123] The way to calculate the weights of training samples in this training process can be referred to the following formula:

[0124]

[0125] Among them, w t+1 (i) w t (i) represents the weight of the training sample; α t represents the classifier weight; yi represents the label of the training sample; f t (x) represents the output value of the weak classifier; z t is the normalization factor of the probability distribution, which ensures that w i+1 (n) Meet the requirements of probability distribution.

[0126] In step S62, multiple weak classifiers are trained based on the training samples and the normalized sample weights. During the training process of the weak classifiers, the classification error rate of the weak classifiers is calculated, and the weights of each weak classifier and the weights of the training samples are calculated and updated, thereby updating the various parameters in the weak classifiers and improving the classification performance of the weak classifiers.

[0127] Step S63, if the classification error rate is less than or equal to the first preset threshold, stop iteration and obtain the target weak classifier;

[0128] In step S63, when the classification error rate is less than or equal to the first preset threshold, it indicates that the classification error rate of the model has been reduced to the expected level, that is, it means that the weak classifier model has been trained, and the iteration can be stopped to obtain the target weak classifier.

[0129] Among them, the first preset threshold can be set according to actual needs. For example, the first preset threshold can be set to 0.005, so that the weak classifier stops iterating after reaching a higher classification accuracy. The first preset threshold can also be set to 0.001, or 0.0001, to further improve the classification accuracy of the weak classifier. The specific value of the first preset threshold is not limited here.

[0130] Step S64, if the classification error rate is greater than or equal to the second preset threshold, stop the iteration, adjust the normalized sample weight, and return to execute the training step until the classification error rate of the weak classifier finally trained is less than or equal to the first preset threshold;

[0131] In step S64, when the classification error rate is greater than or equal to the second preset threshold, it indicates that the classification error rate of the weak classifier is high and the training effect is poor, and it is necessary to stop the iteration and adjust the normalized sample weights, and then re-execute the training steps and start retraining until the classification error rate of the weak classifier obtained by the training is less than or equal to the first preset threshold. This step can effectively avoid the waste of a large amount of time cost and computing resources caused by continuing to train the weak classifier when the training effect is poor.

[0132] It should be noted that the second preset threshold can be set according to actual needs. For example, the second preset threshold can be set to 0.5, so that the weak classifier stops iterative training only when the classification error rate is very high. Of course, the second preset threshold can also be set to other values. The specific value of the second preset threshold is not limited here.

[0133] Step S65, if the classification error rate is the other case, return to execute the training step until the classification error rate of the weak classifier finally trained is less than or equal to the first preset threshold.

[0134] In this embodiment, the weights of the training samples are normalized to help the weak classifier reduce computational complexity and improve computational efficiency; then, multiple weak classifiers are trained based on the training samples and the normalized sample weights; during the training process of the weak classifier, the classification error rate of the weak classifier is calculated, and the weights of each weak classifier and the weights of the training samples are calculated and updated, thereby updating various parameters in the weak classifier and improving the classification performance of the weak classifier; in addition, the training status of the weak classifier model is learned based on the classification error rate of the weak classifier, and different training strategies are adopted for different classification error rates, so that the target weak classifier model finally trained achieves the expected classification performance.

[0135] Step S70, combining the multiple target weak classifiers into the integrated support vector machine model.

[0136] It should be noted that the classification result output by the integrated support vector machine model is:

[0137]

[0138] Wherein, H(x) represents the classification result output by the integrated support vector machine model; sgn(*) represents the sign function; α i represents the weight of the i-th target weak classifier; f i (x i ) represents the i-th target weak classifier in the i-th sample data x i The output value on .

[0139] Referring to the above formula, the classification results of each target weak classifier can be combined to obtain the classification result output by the integrated support vector machine model.

[0140] In the third embodiment of the present application, by making the feature parameters into sample data sets of multiple categories, and dividing them into training subsets and test subsets, and then initializing the weights of the training samples in the training subsets, the production and preprocessing of the data sets corresponding to the feature parameters are completed; by training multiple weak classifiers respectively, and forming an integrated support vector machine model with the target weak classifiers finally obtained through training and whose classification performance meets the expected requirements, the advantages of each target weak classifier are integrated, while reducing the overfitting risk that a single target weak classifier may have, and the weight of the correctly classified samples can be increased to balance the error rate, thereby improving the processing capacity of the unbalanced data set. In addition, by combining multiple weak classifiers, each target weak classifier can focus on different aspects of the data, thereby constructing an integrated support vector machine model with stronger classification performance.

[0141] This application also provides a photovoltaic array fault diagnosis device, please refer to Figure 5 , the photovoltaic array fault diagnosis device comprises:

[0142] The parameter acquisition module 10 is used to acquire characteristic parameters of the photovoltaic array; wherein the characteristic parameters at least include the maximum power point voltage, the maximum power point current and the fill factor;

[0143] The fault diagnosis module 80 is used to perform fault diagnosis on the characteristic parameters of the photovoltaic array based on the maximum power point voltage, the maximum power point current and the fill factor using a pre-trained integrated support vector machine model to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

[0144] The photovoltaic array fault diagnosis device provided by the present application adopts the photovoltaic array fault diagnosis method in the above embodiment, which can solve the technical problem that the existing photovoltaic array fault diagnosis methods mostly use simulation software to establish simulation models, which are difficult to model and lack accuracy, and are difficult to meet the fault diagnosis needs. Compared with the prior art, the beneficial effects of the photovoltaic array fault diagnosis device provided by the present application are the same as the beneficial effects of the photovoltaic array fault diagnosis method provided by the above embodiment, and the other technical features in the photovoltaic array fault diagnosis device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0145] The present application provides a photovoltaic array fault diagnosis device, which includes: a data acquisition module, a data processing and storage module, a fault diagnosis module and a communication module;

[0146] The data acquisition module is used to collect power generation information of the photovoltaic array and transmit the power generation information to the data processing and storage module;

[0147] The data processing and storage module is used to pre-process the power generation information to obtain the characteristic parameters in the photovoltaic array fault diagnosis method in any of the above embodiments;

[0148] The fault diagnosis module is used to store and run a computer program; wherein the computer program is configured to implement the steps of the photovoltaic array fault diagnosis method in any of the above embodiments;

[0149] The communication module is used to transmit the power generation information and the fault diagnosis result to the host computer main control module.

[0150] Reference below Figure 6 , which shows a schematic diagram of the structure of a photovoltaic array fault diagnosis device suitable for implementing the embodiment of the present application. The photovoltaic array fault diagnosis device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The photovoltaic array fault diagnosis device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0151] like Figure 6 As shown, the photovoltaic array fault diagnosis device may include a data acquisition module, a data processing and storage module, a fault diagnosis module and a communication module. Among them, the data acquisition module may be an electrical inductor or an electrical sensor that realizes the corresponding acquisition function, and the data acquisition module is electrically connected to the photovoltaic array to obtain the characteristic parameters of the photovoltaic array; the data processing and storage module and the fault diagnosis module may be modules that integrate processors such as a central processing unit and a graphics processor and memories such as random access memory (RAM: Random Access Memory) and read-only memory (ROM: Read Only Memory); the communication module may allow the photovoltaic array fault diagnosis device to communicate with other devices wirelessly or wired to exchange data. For example, the communication module may enable the photovoltaic array fault diagnosis device to communicate with the host computer main control module through the RS485 bus to exchange data. Although the figure shows a photovoltaic array fault diagnosis device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have alternatively.

[0152] The photovoltaic array fault diagnosis device provided by the present application adopts the photovoltaic array fault diagnosis method in the above embodiment, which can solve the technical problem that the existing photovoltaic array fault diagnosis methods mostly use simulation software to establish simulation models, which has high modeling difficulty, insufficient accuracy, and is difficult to meet the fault diagnosis needs. Compared with the prior art, the beneficial effects of the photovoltaic array fault diagnosis device provided by the present application are the same as the beneficial effects of the photovoltaic array fault diagnosis method provided by the above embodiment, and the other technical features in the photovoltaic array fault diagnosis device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0153] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0154] 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 who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0155] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the photovoltaic array fault diagnosis method in the above-mentioned embodiment.

[0156] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0157] The computer-readable storage medium may be included in the photovoltaic array fault diagnosis device; or may exist independently without being assembled into the photovoltaic array fault diagnosis device.

[0158] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the photovoltaic array fault diagnosis device, the photovoltaic array fault diagnosis device: obtains characteristic parameters of the photovoltaic array; wherein the characteristic parameters include at least the maximum power point voltage, the maximum power point current and the fill factor; based on the maximum power point voltage, the maximum power point current and the fill factor, a pre-trained integrated support vector machine model is used to perform fault diagnosis on the characteristic parameters of the photovoltaic array to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

[0159] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0160] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0161] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0162] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned photovoltaic array fault diagnosis method, and can solve the technical problem that the existing photovoltaic array fault diagnosis method mostly uses simulation software to establish a simulation model, which has high modeling difficulty, insufficient accuracy, and is difficult to meet the fault diagnosis requirements. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as the beneficial effects of the photovoltaic array fault diagnosis method provided by the above-mentioned embodiment, and will not be repeated here.

[0163] The present application also provides a computer program product, including a computer program, wherein when the computer program is executed by a processor, the steps of the photovoltaic array fault diagnosis method as described above are implemented.

[0164] The computer program product provided by the present application can solve the technical problem that the existing photovoltaic array fault diagnosis methods mostly use simulation software to establish simulation models, which are difficult to model and lack accuracy, and are difficult to meet the fault diagnosis needs. Compared with the prior art, the beneficial effects of the computer program product provided by the present application are the same as the beneficial effects of the photovoltaic array fault diagnosis method provided by the above embodiment, and will not be repeated here.

[0165] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A photovoltaic array fault diagnosis method, characterized in that: The method comprises: Acquiring characteristic parameters of a photovoltaic array; wherein the characteristic parameters include at least a maximum power point voltage, a maximum power point current and a fill factor; Based on the maximum power point voltage, the maximum power point current and the fill factor, a pre-trained integrated support vector machine model is used to perform fault diagnosis on characteristic parameters of the photovoltaic array to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

2. The method according to claim 1, characterized in that After the step of obtaining characteristic parameters of the photovoltaic array, the method further comprises: Normalizing the maximum power point voltage and the maximum power point current to obtain a normalized maximum power point voltage and a normalized maximum power point current; The normalized maximum power point voltage and the normalized maximum power point current are added to the characteristic parameters.

3. The method according to claim 1 or 2, characterized in that Before the step of performing fault diagnosis on characteristic parameters of the photovoltaic array using a pre-trained integrated support vector machine model based on the maximum power point voltage, the maximum power point current and the fill factor to obtain a fault diagnosis result of the photovoltaic array, the method comprises: Based on the characteristic parameters, preparing sample data sets of multiple categories, and dividing the sample data sets into a training subset and a test subset; Initializing the weights of the training samples in the training subset to obtain initialized sample weights; Based on the training subset and the initialization sample weight, multiple weak classifiers are trained respectively to obtain multiple target weak classifiers; wherein the weak classifier is a model based on a support vector machine; A plurality of the target weak classifiers are combined into the integrated support vector machine model.

4. The method according to claim 3, characterized in that The step of training a plurality of weak classifiers based on the training subset and the initialization sample weights to obtain a plurality of target weak classifiers comprises: Based on the sample distribution, n training samples are selected from the training subset, and the weights of the training samples are normalized to obtain normalized sample weights; A training step, based on the training samples and the normalized sample weights, training a plurality of the weak classifiers according to a preset number of training rounds; during the training process of the weak classifiers, calculating the classification error rate of the weak classifiers, and calculating and updating the weight of each of the weak classifiers and the weight of the training samples; If the classification error rate is less than or equal to a first preset threshold, the iteration is stopped to obtain the target weak classifier; If the classification error rate is greater than or equal to the second preset threshold, stop the iteration, adjust the normalized sample weight, and return to execute the training step until the classification error rate of the weak classifier finally trained is less than or equal to the first preset threshold; If the classification error rate is any other case, return to execute the training step until the classification error rate of the weak classifier finally obtained by training is less than or equal to the first preset threshold.

5. The method according to claim 3, characterized in that The parameters of the support vector machine in the weak classifier include at least a kernel function, and the kernel function is set to a Gaussian kernel function.

6. The method according to claim 3, characterized in that The classification result output by the integrated support vector machine model is: Wherein, H(x) represents the classification result output by the integrated support vector machine model; sgn(*) represents the sign function; α i represents the weight of the i-th target weak classifier; f i (x i ) represents the i-th target weak classifier in the i-th sample data x i The output value on .

7. A photovoltaic array fault diagnosis device, characterized in that: The device comprises: A parameter acquisition module, used to acquire characteristic parameters of the photovoltaic array; wherein the characteristic parameters at least include maximum power point voltage, maximum power point current and fill factor; A fault diagnosis module is used to perform fault diagnosis on characteristic parameters of the photovoltaic array based on the maximum power point voltage, the maximum power point current and the fill factor using a pre-trained integrated support vector machine model to obtain a fault diagnosis result of the photovoltaic array; wherein the integrated support vector machine model is obtained based on support vector machine combined with integrated learning training.

8. A photovoltaic array fault diagnosis device, characterized in that: The device comprises: a data acquisition module, a data processing and storage module, a fault diagnosis module and a communication module; The data acquisition module is used to collect power generation information of the photovoltaic array and transmit the power generation information to the data processing and storage module; The data processing and storage module is used to pre-process the power generation information to obtain the characteristic parameters in the photovoltaic array fault diagnosis method according to any one of claims 1 to 6; The fault diagnosis module is used to store and run a computer program; wherein the computer program is configured to implement the steps of the photovoltaic array fault diagnosis method according to any one of claims 1 to 6; The communication module is used to transmit the power generation information and the fault diagnosis result to the host computer main control module.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the photovoltaic array fault diagnosis method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the photovoltaic array fault diagnosis method according to any one of claims 1 to 6 are implemented.

Images