Method and device for determining fault detection model of photovoltaic system, and storage medium

Abstract

The invention discloses a method and a device for determining a fault detection model of a photovoltaic system, a fault detection method of the photovoltaic system, a storage medium, an electronic device and a computer program product. The method comprises the following steps: acquiring historical operation data of the photovoltaic system to be detected; and adopting a convolutional neural network model to extract fault features in the historical operation data to obtain feature data. According to the feature data, a bidirectional gating circulation unit model is constructed, and a bidirectional gating circulation unit is used for outputting a prediction result based on forward data and backward data of the feature data in the time sequence. And training the bidirectional gating cycle unit model by adopting pre-marked training data so as to determine a fault detection model of the photovoltaic system. The model can determine the current state of the photovoltaic system based on the historical information and future information of the operation data, so that the output fault detection result is more accurate.

Description

Technical Field

[0001] The present application relates to the photovoltaic field, and specifically, to a method and device for determining a fault detection model of a photovoltaic system, a fault detection method of a photovoltaic system, a storage medium, an electronic device, and a computer program product. Background Art

[0002] As the global demand for renewable energy continues to increase, photovoltaic energy, as an important renewable energy source, has been widely used around the world. The generation of photovoltaic energy depends on photovoltaic systems. However, photovoltaic systems may encounter various faults in actual operation, which will not only reduce the power generation efficiency of photovoltaic systems, but also may cause serious safety hazards.

[0003] In the related art, traditional physical models and statistical analysis methods are mainly used to perform fault detection of photovoltaic systems.

[0004] However, the fault detection methods in related technologies usually require a lot of prior knowledge and experience and are not very accurate. Summary of the invention

[0005] Embodiments of the present application provide a method and device for determining a fault detection model for a photovoltaic system, a fault detection method for a photovoltaic system, a storage medium, an electronic device, and a computer program product.

[0006] According to one aspect of an embodiment of the present application, a method for determining a fault detection model of a photovoltaic system is provided, the method comprising: acquiring historical operation data of a photovoltaic system to be detected; extracting fault features in the historical operation data using a convolutional neural network model to obtain feature data, wherein the feature data is used to indicate a fault condition of the photovoltaic system; constructing a bidirectional gated recurrent unit model based on the feature data, wherein the bidirectional gated recurrent unit is used to output a prediction result based on forward data and backward data of the feature data in a time series, the fault detection model comprising a convolutional neural network model and the bidirectional gated recurrent unit model; and training the bidirectional gated recurrent unit model using pre-labeled training data to determine a fault detection model of the photovoltaic system, wherein the fault detection model of the photovoltaic system is used to detect a fault condition of the photovoltaic system.

[0007] In an exemplary embodiment, constructing a bidirectional gated recurrent unit model based on the feature data includes: using the feature data to perform time series modeling to construct the bidirectional gated recurrent unit model, wherein the bidirectional gated recurrent unit includes a forward gated recurrent unit and a backward gated recurrent unit, the forward gated recurrent unit is used to process the feature data in the positive order of the time series, and the backward gated recurrent unit is used to process the feature data in the reverse order of the time series; wherein the output layer of the bidirectional gated recurrent unit model is respectively connected to the output of the forward gated recurrent unit and the output of the backward gated recurrent unit, and the output layer is used to combine the output of the forward gated recurrent unit and the output of the backward gated recurrent unit for output.

[0008] In an exemplary embodiment, the use of a convolutional neural network model to extract fault features in the historical operation data to obtain feature data includes: filtering the historical operation data to extract fault features in the historical operation data and generating a feature map of the historical operation data; and converting the feature map into a feature vector as the feature data.

[0009] In an exemplary embodiment, before filtering the historical operation data, extracting fault features in the historical operation data, and generating a feature map of the historical operation data, the method for determining the fault detection model of the photovoltaic system further includes: preprocessing the historical operation data, wherein the preprocessing includes at least one of the following: data cleaning and normalization.

[0010] In an exemplary embodiment, the method of training the bidirectional gated cyclic unit model with pre-labeled training data to determine the fault detection model of the photovoltaic system includes: acquiring pre-labeled training data, wherein the training data includes fault data and normal data; training the bidirectional gated cyclic unit model with the training data until the prediction accuracy of the bidirectional gated cyclic unit model meets the standard; and recording and saving the parameters of the bidirectional gated cyclic unit model that meets the prediction accuracy to determine the fault detection model of the photovoltaic system.

[0011] According to another aspect of an embodiment of the present application, a method for fault detection of a photovoltaic system is also provided. The method applies a fault detection model of a photovoltaic system as described above, and the method includes: acquiring operating data of the photovoltaic system to be detected in real time; inputting the operating data into the fault detection model to determine a fault detection result of the photovoltaic system.

[0012] According to another aspect of an embodiment of the present application, a device for determining a fault detection model of a photovoltaic system is also provided, the device comprising:

[0013] A data acquisition module, used to acquire historical operation data of the photovoltaic system to be detected;

[0014] A feature extraction module, used to extract fault features in the historical operation data using a convolutional neural network model to obtain feature data, wherein the feature data is used to indicate a fault condition of the photovoltaic system;

[0015] A unit model construction module, used to construct a bidirectional gated recurrent unit model according to the feature data, wherein the bidirectional gated recurrent unit is used to output a prediction result based on forward data and backward data of the feature data in a time series, and the fault detection model includes a convolutional neural network model and the bidirectional gated recurrent unit model;

[0016] The model determination module is used to train the bidirectional gated cyclic unit model using pre-labeled training data to determine the fault detection model of the photovoltaic system.

[0017] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the method for determining the fault detection model of the photovoltaic system when running.

[0018] According to another aspect of an embodiment of the present application, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the method for determining the fault detection model of the photovoltaic system through the computer program.

[0019] According to another aspect of the embodiments of the present application, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps of the method described in each embodiment of the present application are implemented.

[0020] The method for determining the fault detection model of the above-mentioned photovoltaic system first obtains the historical operation data of the photovoltaic system to be detected. Then, a convolutional neural network model is used to extract the fault features in the historical operation data to obtain feature data. Next, a model is built on the time series based on the feature data to construct a bidirectional gated recurrent unit model. The bidirectional gated recurrent unit is used to output the prediction result based on the forward data and backward data of the feature data on the time series, so that the output prediction result contains past and future information, and the prediction result is more accurate. The bidirectional gated recurrent unit model is trained using pre-labeled training data to determine the fault detection model of the photovoltaic system. Since the fault detection model includes a bidirectional gated recurrent unit model, the current state of the photovoltaic system can be predicted based on the historical information and future information of the operation data, so the fault detection result of the photovoltaic system obtained using the fault detection model is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] 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.

[0022] 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.

[0023] Figure 1 is a hardware structure block diagram of a method for determining a fault detection model of a photovoltaic system according to an embodiment of the present application;

[0024] Figure 2 is a flow chart of a method for determining a fault detection model of a photovoltaic system according to an embodiment of the present application;

[0025] Figure 3 is a second flowchart of a method for determining a fault detection model of a photovoltaic system according to an embodiment of the present application;

[0026] Figure 4 is a flowchart of a method for determining a fault detection model of a photovoltaic system according to an embodiment of the present application;

[0027] Figure 5 is a fourth flow chart of a method for determining a fault detection model of a photovoltaic system according to an embodiment of the present application;

[0028] Figure 6 is a flow chart of a photovoltaic system fault detection method according to an embodiment of the present application;

[0029] Figure 7 It is a comprehensive flow chart of a photovoltaic system fault detection method and a photovoltaic system fault detection model determination method according to an embodiment of the present application;

[0030] Figure 8 It is a structural block diagram of a device for determining a fault detection model of a photovoltaic system according to an embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, the process, method, system, product or equipment comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0033] The method embodiments provided in the embodiments of the present application can be executed in a computer terminal or a similar computing device. Taking running on a computer terminal as an example, Figure 1 FIG. 1 is a hardware structure block diagram of a computer terminal of a method for determining a fault detection model of a photovoltaic system according to an embodiment of the present application. Figure 1 As shown, the computer terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a microprocessor (Microprocessor Unit, referred to as MPU) or a programmable logic device (Programmable logic device, referred to as PLD)) and a memory 104 for storing data. In an exemplary embodiment, the computer terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1The structure shown is only for illustration and does not limit the structure of the above-mentioned computer terminal. Figure 1 More or fewer components as shown, or with Figure 1 Equivalent functions or comparisons shown Figure 1 A different configuration with more features is shown.

[0034] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the method for determining the fault detection model of the photovoltaic system in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0035] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of a computer terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0036] In this embodiment, a method for determining a fault detection model of a photovoltaic system is provided. Figure 2 4 is a flow chart of an optional method for determining a fault detection model of a photovoltaic system according to an embodiment of the present application, the flow chart comprising the following steps S200-S230:

[0037] Step S200, obtaining historical operation data of the photovoltaic system to be detected.

[0038] Among them, the historical operation data includes the temperature of the photovoltaic modules (photovoltaic panels) of the photovoltaic system, solar irradiance and other information. This information changes over time and can be represented in the form of a time series.

[0039] Step S210, using a convolutional neural network model to extract fault features in historical operation data to obtain feature data.

[0040] The characteristic data is used to indicate the fault condition of the photovoltaic system. The characteristic data may be abnormal data in the historical operation data, such as data showing that the temperature of the photovoltaic module (photovoltaic panel) is too high or too low, indicating that the photovoltaic module may have a fault.

[0041] Step S220: construct a bidirectional gated recurrent unit model according to the feature data.

[0042] The bidirectional gated recurrent unit is used to output a prediction result based on the forward data and the backward data of the feature data in the time series, and the fault detection model includes a convolutional neural network model and a bidirectional gated recurrent unit model;

[0043] Step S230 , using pre-labeled training data to train the bidirectional gated recurrent unit model to determine a fault detection model for the photovoltaic system.

[0044] Among them, the fault detection model of the photovoltaic system is used to detect the fault conditions of the photovoltaic system.

[0045] In this embodiment, the historical operation data of the photovoltaic system to be detected is first obtained. Then, a convolutional neural network model is used to extract fault features in the historical operation data to obtain feature data. Next, a model is built on the time series based on the feature data to construct a bidirectional gated recurrent unit model. The bidirectional gated recurrent unit is used to output a prediction result based on the forward data and backward data of the feature data on the time series, so that the output prediction result contains past and future information, and the prediction result is more accurate. The bidirectional gated recurrent unit model is trained using pre-labeled training data to determine a fault detection model for the photovoltaic system. Since the fault detection model includes a bidirectional gated recurrent unit model, the current state of the photovoltaic system can be predicted based on the historical information and future information of the operation data, so the fault detection result of the photovoltaic system obtained using the fault detection model is more accurate.

[0046] In one embodiment, step S220 constructs a bidirectional gated recurrent unit model based on the feature data, including:

[0047] The feature data is used for time series modeling to construct a bidirectional gated recurrent unit model.

[0048] Among them, the bidirectional gated recurrent unit (BI GRU) includes a forward gated recurrent unit and a backward gated recurrent unit. The forward gated recurrent unit is used to process the feature data in the positive order of the time series, and the backward gated recurrent unit is used to process the feature data in the reverse order of the time series.

[0049] The unidirectional gated recurrent unit (GRU) consists of two gates: a reset gate and an update gate, which are used to regulate information transmission. The update gate determines the current hidden layer state h. t How many previous hidden layer states are retained in t-1 ) information. It processes h through the activation function t-1 and the current input x t , the activation result can be obtained, the expression is as follows:

[0050] z t =σ(W zx x t +W zh h t-1 +b z )

[0051] Among them, z t is the activation result of the update gate, representing the current hidden layer state h t How many previous hidden layer states h are retained t-1 information, σ is the activation function, W zx For input x t The associated weight matrix, W zh is the same as the previous hidden layer state h t-1 The information-related weight matrix, b z is the bias term.

[0052] Specifically, update the activation value z of the gate t The calculation process is to first pass the weight matrix W zx and W zh and the bias term b z For input x t and the previous hidden state h t-1 Perform a linear transformation and then compress it through the sigmoid activation function. The result of the compression process is the activation value z of the update gate t . Update the gate activation value z t When it is close to 1, it indicates the hidden state h at the current moment. t More hidden states h from the previous moment should be retained t-1 information; and when z t When it is close to 0, it indicates the hidden state h at the current moment t The hidden state h of the previous moment should be kept less t-1 That is, most of the information should come from the current input x t ,By updating the gate, information can be selectively transferred between different time steps of the sequence.

[0053] The reset gate determines whether to write candidate memory states from the previous time. Similar to the update gate, the reset gate processes h through the activation function t-1 and x t , a larger activation result means more information is written to The reset gate is defined as follows:

[0054] r t =σ(W rx x t +W rh h t-1 +b r )

[0055] Among them, r t is the activation result of the reset gate, representing the current hidden layer state h t How many previous hidden layer states h are written t-1 information, σ is the activation function, W rx For input x t The associated weight matrix, W rh is the same as the previous hidden layer state h t-1 The information-related weight matrix, b r is the bias term.

[0056] GRU combines the reset gate r t With h t-1 and x t , the candidate memory state is formed by the following expression

[0057]

[0058] The current hidden layer state (h t ) is expressed as follows: t-1 ) and candidate memory states Combined to get:

[0059]

[0060] Among them, x t Indicates the input at the current moment; h t-1 and h t Represent the hidden layer states at the previous and current moments respectively; Represents the candidate memory state; r t and z t Represent the reset gate and update gate respectively; W zx , W rx and are the weight matrices related to the update gate, reset gate, and candidate memory state; W zh , W rh and h related to the update gate, reset gate and candidate memory state respectively t-1 The weight matrix of z 、b r and They represent the corresponding bias terms respectively.

[0061] Among them, the output layer of the bidirectional gated recurrent unit model is connected to the output of the forward gated recurrent unit and the output of the backward gated recurrent unit respectively, and the output layer is used to combine the output of the forward gated recurrent unit and the output of the backward gated recurrent unit for output.

[0062] Specifically, the bidirectional gated recurrent unit is composed of two forward gated recurrent units and backward gated recurrent units (two hidden layers) with different information transmission directions, which can better extract input information. The current state is not only affected by historical data, but also by future information. The output of the forward gated recurrent unit and the backward gated recurrent unit are connected to the same output layer, and the output of the current state (y t ) is expressed as follows:

[0063] y t =[h >> ,h << ]

[0064] Among them, h >> ,h << They represent the outputs of the forward gated recurrent unit and the backward gated recurrent unit, respectively.

[0065] In this embodiment, a bidirectional gated cyclic unit model is constructed, which can predict the current state based on the historical information and future information of the operating data, and the obtained prediction result is more accurate. Therefore, by using the bidirectional gated cyclic unit model as part of the fault detection model, the fault detection result of the photovoltaic system is also more accurate.

[0066] In one embodiment, Figure 3 As shown, step S210 uses a convolutional neural network model to extract fault features in historical operation data to obtain feature data. It includes: steps S300-S310. Among them:

[0067] Step S300 , filtering the historical operation data, extracting fault features in the historical operation data, and generating a feature map of the historical operation data.

[0068] Step S310, converting the feature map into a feature vector as feature data.

[0069] Specifically, convolutional neural networks (CNNs) are used to extract features from historical operation data and enhance feature representation capabilities. CNNs contain input layers, convolutional layers, and pooling layers, which are stacked layer by layer in the network as feature extractors. The input historical operation data passes through a series of filters to generate feature maps.

[0070] The generated convolutional feature map is then flattened and connected into a feature vector so that it can be processed by the fully connected layer, and then the feature vector is input into the fully connected layer for further feature combination. Finally, the output result of the fully connected layer is classified through the output layer to obtain the probability of the feature data belonging to each category, thus realizing feature extraction.

[0071] In this embodiment, a convolutional neural network model is used to extract fault features from historical operation data, and characteristic data is obtained to facilitate subsequent identification of fault features.

[0072] In one embodiment, Figure 4 As shown, in step S300, filtering is performed on the historical operation data, fault features in the historical operation data are extracted, and a feature map of the historical operation data is generated. Previously, the method for determining the fault detection model of the photovoltaic system also includes: step S400, preprocessing the historical operation data.

[0073] The preprocessing includes at least one of the following: data cleaning and normalization. Cleaning the historical operation data can remove outliers or noise. Normalizing the historical operation data makes it more suitable for neural network processing.

[0074] In this embodiment, the data is preprocessed to remove interference information in the data, so that the subsequent processing results of the neural network are more accurate.

[0075] In one embodiment, Figure 5 As shown, step S230, using pre-labeled training data to train the bidirectional gated recurrent unit model to determine the fault detection model of the photovoltaic system. It includes: steps S500-S520. Among them:

[0076] Step S500, obtaining pre-labeled training data.

[0077] The training data includes fault data and normal data.

[0078] Step S510, using training data to train the bidirectional gated recurrent unit model until the prediction accuracy of the bidirectional gated recurrent unit model meets the standard.

[0079] This can be done by adjusting model parameters to minimize prediction errors through optimization algorithms such as gradient descent. The model’s performance is then verified on an independent test dataset to determine whether the model’s prediction accuracy meets the target.

[0080] Step S520 , recording and saving the parameters of the bidirectional gated cyclic unit model that meets the prediction accuracy standard, so as to determine the fault detection model of the photovoltaic system.

[0081] Specifically, the parameters and structure of the trained fault detection model are stored so that the fault detection model can be subsequently applied to perform real-time fault detection on the photovoltaic system.

[0082] In this embodiment, after the fault detection model is constructed, the fault detection model is trained using training data, and the accuracy of the fault detection model is verified and evaluated. Finally, the trained fault detection model is stored and recorded so that the fault detection model can be used to perform fault detection on the photovoltaic system and improve the accuracy of fault detection.

[0083] In one embodiment, Figure 6 , provides a photovoltaic system fault detection method, the method applies the photovoltaic system fault detection model in the above embodiment, the method includes: steps S600-S610.

[0084] Step S600: acquiring the operating data of the photovoltaic system to be detected in real time.

[0085] Specifically, after the operation data of the photovoltaic system to be detected is acquired, the operation data is preprocessed, such as data cleaning and normalization.

[0086] Step S610: inputting the operation data into a fault detection model to determine a fault detection result of the photovoltaic system.

[0087] For example, a schematic diagram of combining a photovoltaic system fault detection method with a photovoltaic system fault detection model determination method is shown in FIG. Figure 7 As shown, the scheme includes an offline phase and an online phase. The fault detection model of the photovoltaic system is determined in the offline phase, and the fault detection model of the photovoltaic system is used to perform fault detection in the online phase.

[0088] Specifically, the offline phase includes:

[0089] S101, data collection and preprocessing, which includes steps S200 and S400 in the above embodiment.

[0090] S102, feature extraction, includes step S210 in the above embodiment.

[0091] S103, model training, includes steps S220-S230 in the above embodiment.

[0092] S104, model storage, which includes step S520 in the above embodiment.

[0093] The online phase includes:

[0094] S201, real-time monitoring and data updating, which includes step S600 in the present embodiment.

[0095] S202, fault detection, includes step S610 in the present embodiment.

[0096] S203, result output: that is, the result obtained by fault detection is output to an external host computer.

[0097] In this embodiment, the operation data of the photovoltaic system is collected in real time, and the operation data is input into the fault detection model for fault detection and diagnosis, and the detection results of the output model include the fault type and possible fault location. This method can significantly improve the accuracy and speed of fault detection, and is suitable for the operation and maintenance management of large-scale photovoltaic power stations, especially in emergency situations that require rapid response, and can quickly locate faults to avoid system shutdown or damage.

[0098] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0099] In this embodiment, a device for determining a fault detection model of a photovoltaic system is also provided, and the device for determining a fault detection model of a photovoltaic system is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0100] Figure 8 is a structural block diagram of an optional photovoltaic system fault detection model determination device according to an embodiment of the present application; Figure 8 As shown, including:

[0101] The data acquisition module 801 is used to acquire the historical operation data of the photovoltaic system to be detected.

[0102] The feature extraction module 802 is used to extract fault features in the historical operation data using a convolutional neural network model to obtain feature data, wherein the feature data is used to indicate the fault condition of the photovoltaic system.

[0103] The unit model construction module 803 is used to construct a bidirectional gated recurrent unit model according to the feature data, wherein the bidirectional gated recurrent unit is used to output a prediction result based on the forward data and backward data of the feature data in a time series, and the fault detection model includes a convolutional neural network model and a bidirectional gated recurrent unit model.

[0104] The model determination module 804 is used to train the bidirectional gated recurrent unit model using pre-labeled training data to determine a fault detection model for the photovoltaic system.

[0105] Through the above-mentioned device, the historical operation data of the photovoltaic system to be detected is first obtained. Then, a convolutional neural network model is used to extract fault features in the historical operation data to obtain feature data. Next, a model is built on the time series based on the feature data to construct a bidirectional gated recurrent unit model. The bidirectional gated recurrent unit is used to output a prediction result based on the forward data and backward data of the feature data on the time series, so that the output prediction result contains past and future information, and the prediction result is more accurate. The bidirectional gated recurrent unit model is trained using pre-labeled training data to determine the fault detection model of the photovoltaic system. Since the fault detection model includes a bidirectional gated recurrent unit model, the current state of the photovoltaic system can be predicted based on the historical information and future information of the operation data, so the fault detection result of the photovoltaic system obtained using the fault detection model is more accurate.

[0106] In an exemplary embodiment, the unit model construction module 803 is also used to use feature data to perform time series modeling to construct a bidirectional gated recurrent unit model, wherein the bidirectional gated recurrent unit includes a forward gated recurrent unit and a backward gated recurrent unit, the forward gated recurrent unit is used to process the feature data in the positive order of the time series, and the backward gated recurrent unit is used to process the feature data in the reverse order of the time series; wherein the output layer of the bidirectional gated recurrent unit model is respectively connected to the output of the forward gated recurrent unit and the output of the backward gated recurrent unit, and the output layer is used to combine the output of the forward gated recurrent unit and the output of the backward gated recurrent unit for output.

[0107] In an exemplary embodiment, the feature extraction module 802 is further used to filter the historical operation data, extract fault features in the historical operation data, generate a feature map of the historical operation data, and convert the feature map into a feature vector as feature data.

[0108] In an exemplary embodiment, the model determination module 804 is also used to obtain pre-labeled training data, wherein the training data includes fault data and normal data; the bidirectional gated cyclic unit model is trained using the training data until the prediction accuracy of the bidirectional gated cyclic unit model meets the standard; and the parameters of the bidirectional gated cyclic unit model that meets the prediction accuracy are recorded and saved to determine the fault detection model of the photovoltaic system.

[0109] An embodiment of the present application further provides a storage medium, which includes a stored program, wherein the program executes any of the above methods when it is run.

[0110] Optionally, in this embodiment, the storage medium may be configured to store program codes for executing the following steps:

[0111] S1, obtaining historical operation data of the photovoltaic system to be detected.

[0112] S2, using a convolutional neural network model to extract fault features in historical operation data to obtain feature data, wherein the feature data is used to indicate a fault condition of the photovoltaic system.

[0113] S3, constructing a bidirectional gated recurrent unit model according to the feature data, wherein the bidirectional gated recurrent unit is used to output a prediction result based on the forward data and backward data of the feature data in a time series, and the fault detection model includes a convolutional neural network model and a bidirectional gated recurrent unit model.

[0114] S4, using pre-labeled training data to train the bidirectional gated recurrent unit model to determine a fault detection model for the photovoltaic system, wherein the fault detection model for the photovoltaic system is used to detect fault conditions of the photovoltaic system.

[0115] An embodiment of the present application further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0116] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0117] Optionally, in this embodiment, the processor may be configured to perform the following steps through a computer program:

[0118] S1, obtaining historical operation data of the photovoltaic system to be detected.

[0119] S2, using a convolutional neural network model to extract fault features in historical operation data to obtain feature data, wherein the feature data is used to indicate a fault condition of the photovoltaic system.

[0120] S3, constructing a bidirectional gated recurrent unit model according to the feature data, wherein the bidirectional gated recurrent unit is used to output a prediction result based on the forward data and backward data of the feature data in a time series, and the fault detection model includes a convolutional neural network model and a bidirectional gated recurrent unit model.

[0121] S4, using pre-labeled training data to train the bidirectional gated recurrent unit model to determine a fault detection model for the photovoltaic system, wherein the fault detection model for the photovoltaic system is used to detect fault conditions of the photovoltaic system.

[0122] Optionally, in this embodiment, the above-mentioned storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store program codes.

[0123] An embodiment of the present application further provides a computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores the computer program product, and when the computer program is executed by a processor, the steps of the method in each embodiment of the present application are implemented.

[0124] Optionally, in this embodiment, the above computer program may be configured to implement the following steps when executed by a processor:

[0125] S1, obtaining historical operation data of the photovoltaic system to be detected.

[0126] S2, using a convolutional neural network model to extract fault features in historical operation data to obtain feature data, wherein the feature data is used to indicate a fault condition of the photovoltaic system.

[0127] S3, constructing a bidirectional gated recurrent unit model according to the feature data, wherein the bidirectional gated recurrent unit is used to output a prediction result based on the forward data and backward data of the feature data in a time series, and the fault detection model includes a convolutional neural network model and a bidirectional gated recurrent unit model.

[0128] S4, using pre-labeled training data to train the bidirectional gated recurrent unit model to determine a fault detection model for the photovoltaic system, wherein the fault detection model for the photovoltaic system is used to detect fault conditions of the photovoltaic system.

[0129] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.

[0130] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0131] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for determining a fault detection model for a photovoltaic system, characterized in that: The method comprises: Obtain historical operating data of the photovoltaic system to be tested; A convolutional neural network model is used to extract fault features in the historical operation data to obtain feature data, wherein the feature data is used to indicate a fault condition of the photovoltaic system; According to the feature data, a bidirectional gated recurrent unit model is constructed, wherein the bidirectional gated recurrent unit is used to output a prediction result based on forward data and backward data of the feature data in a time series, and the fault detection model includes a convolutional neural network model and the bidirectional gated recurrent unit model; The bidirectional gated recurrent unit model is trained using pre-labeled training data to determine a fault detection model for the photovoltaic system, wherein the fault detection model for the photovoltaic system is used to detect a fault condition of the photovoltaic system.

2. The method according to claim 1, characterized in that The step of constructing a bidirectional gated recurrent unit model according to the characteristic data includes: Using the feature data to perform time series modeling to construct the bidirectional gated recurrent unit model, wherein the bidirectional gated recurrent unit includes a forward gated recurrent unit and a backward gated recurrent unit, the forward gated recurrent unit is used to process the feature data in the positive order of the time series, and the backward gated recurrent unit is used to process the feature data in the reverse order of the time series; Among them, the output layer of the bidirectional gated recurrent unit model is connected to the output of the forward gated recurrent unit and the output of the backward gated recurrent unit respectively, and the output layer is used to combine the output of the forward gated recurrent unit and the output of the backward gated recurrent unit for output.

3. The method according to claim 1, characterized in that The convolutional neural network model is used to extract fault features in the historical operation data to obtain feature data, including: Performing filtering processing on the historical operation data, extracting fault features in the historical operation data, and generating a feature map of the historical operation data; The feature map is converted into a feature vector as the feature data.

4. The method according to claim 3, characterized in that Before filtering the historical operation data, extracting fault features from the historical operation data, and generating a feature map of the historical operation data, the method further includes: The historical operation data is preprocessed, wherein the preprocessing includes at least one of the following: data cleaning and normalization.

5. The method according to claim 1, characterized in that The method of training the bidirectional gated recurrent unit model using pre-labeled training data to determine a fault detection model for the photovoltaic system includes: Acquire pre-labeled training data, wherein the training data includes fault data and normal data; Using the training data to train the bidirectional gated recurrent unit model until the prediction accuracy of the bidirectional gated recurrent unit model meets the standard; The parameters of the bidirectional gated cyclic unit model that meets the prediction accuracy standard are recorded and saved to determine the fault detection model of the photovoltaic system.

6. A photovoltaic system fault detection method, characterized in that: The method applies the fault detection model of the photovoltaic system according to any one of claims 1 to 5, and the method comprises: Obtain the operating data of the photovoltaic system to be tested in real time; The operating data is input into the fault detection model to determine a fault detection result of the photovoltaic system.

7. A device for determining a fault detection model of a photovoltaic system, characterized in that: The device comprises: A data acquisition module, used to acquire historical operation data of the photovoltaic system to be detected; A feature extraction module, used to extract fault features in the historical operation data using a convolutional neural network model to obtain feature data, wherein the feature data is used to indicate a fault condition of the photovoltaic system; A unit model construction module, used to construct a bidirectional gated recurrent unit model according to the feature data, wherein the bidirectional gated recurrent unit is used to output a prediction result based on forward data and backward data of the feature data in a time series, and the fault detection model includes a convolutional neural network model and the bidirectional gated recurrent unit model; The model determination module is used to train the bidirectional gated cyclic unit model using pre-labeled training data to determine the fault detection model of the photovoltaic system.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein the program executes the method according to any one of claims 1 to 6 when executed.

9. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to execute the method according to any one of claims 1 to 6 through the computer program.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images