A photovoltaic power station fault prediction method, system, device and medium
By constructing a fault prediction model based on feature processing and feature fusion, the problem of inaccurate fault prediction in photovoltaic power plants was solved, enabling accurate fault identification and early warning, and improving operation and maintenance efficiency as well as power generation efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 华能(嘉峪关)新能源有限公司
- Filing Date
- 2024-11-30
- Publication Date
- 2026-06-02
AI Technical Summary
Existing methods for predicting faults in photovoltaic power plants are not accurate enough, making it difficult to accurately identify fault types and locations. They are also computationally expensive and have poor adaptability.
A fault prediction model is constructed using algorithms such as convolutional neural networks, multilayer perceptrons, and support vector machines. Through feature processing and feature fusion, faults in photovoltaic electronic equipment are identified, and early warning classification and automatic alarm mechanisms are set up.
It improves the accuracy of fault diagnosis and operation and maintenance efficiency, reduces downtime due to faults, ensures stable operation of photovoltaic systems, and improves power generation efficiency and economic benefits.
Smart Images
Figure CN122137340A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of photovoltaic power generation technology, and specifically to a method, system, equipment and medium for predicting faults in photovoltaic power plants. Background Technology
[0002] The extraction, transportation, and use of fossil fuels generate large amounts of pollutants and greenhouse gases, causing severe environmental damage. Photovoltaic power generation is a clean and pollution-free energy production method, playing a significant role in reducing carbon emissions and mitigating climate change. With increasing global awareness of environmental protection, the demand for clean energy is also growing, providing ample room for the development of photovoltaic power plants.
[0003] When photovoltaic (PV) power generation equipment experiences a major malfunction, it can lead to large-scale load losses in the power system and significantly impact grid stability. Generally, the larger the installed PV capacity, the greater the impact and losses on the grid caused by its malfunctions and power outages. Therefore, real-time monitoring of the electrical and meteorological parameters of PV power generation equipment, predicting the malfunction risk based on these parameters, and then implementing grid emergency measures according to the prediction results can effectively avoid the impact of PV power generation equipment malfunctions on the grid, thereby significantly improving the reliability and economy of the power system.
[0004] The physical failure modeling method predicts system degradation and remaining lifespan by establishing physical failure models for each component in a power electronic circuit. The basic principle of this method is based on in-depth research into component failure mechanisms. By analyzing the stress and damage accumulation of components under different operating conditions, a mathematical model is established to describe the lifespan distribution of the components. This theoretical analysis of the system's failure mechanism provides a relatively accurate basis for fault prediction. However, this method also has some drawbacks. First, it is difficult to obtain an accurate physical model characterizing system degradation or remaining lifespan. Power electronic circuits contain a wide variety of components with complex operating conditions, requiring extensive experimental data and specialized knowledge to establish an accurate physical model. Second, even if a physical model is established, its accuracy may be affected by the uncertainty and complexity of actual operating conditions. Furthermore, the establishment and solution of physical models typically require high computational and time costs. Additionally, if the system's operating conditions change, the physical model may need to be re-established or adjusted, further increasing the difficulty and complexity of fault prediction. Therefore, for... Summary of the Invention The technical problem to be solved by the present invention is to provide a method, system, equipment and medium for predicting faults in photovoltaic power plants, in order to address the shortcomings of the prior art and solve the technical problem of inaccurate fault prediction methods for electronic equipment in current photovoltaic power plants.
[0005] The objective of this invention is achieved through the following technical solutions: A method for predicting faults in a photovoltaic power plant, comprising: Acquire the operating data of the photovoltaic electronic equipment during the current set time period. The operating data includes at least the operating current or voltage data, the power generation of the photovoltaic module, environmental data, and the inverter operating status data. The operating data of each electronic device is input into the trained fault prediction model. The fault prediction model is then used to perform fault analysis on the input feature parameters to obtain the fault prediction results. The fault prediction results are classified for early warning. When the fault prediction results reach the preset alarm threshold, an automatic alarm mechanism is triggered. The fault prediction model includes a feature processing module and a fault identification module. The feature processing module is used to process the input features to obtain feature vectors. The fault identification module is used to perform feature fusion based on the feature vectors and to identify faults based on the fused feature vectors to obtain fault prediction results.
[0006] As a further improvement of the present invention, after acquiring the operating data during the current operation of the photovoltaic electronic device, the method further includes preprocessing the operating data based on historical datasets. The preprocessing includes denoising the acquired operating data and normalizing the denoised operating data. The normalization process includes:
[0007] In the formula, X represents the currently running data. The minimum value among the running data in the historical dataset. This represents the maximum value in the historical dataset.
[0008] As a further improvement of the present invention, the feature processing module employs a convolutional neural network, a multilayer perceptron, or a recurrent neural network.
[0009] As a further improvement of the present invention, the feature processing module employs a multilayer perceptron algorithm and a support vector machine algorithm; The multilayer perceptron algorithm includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is the same as the feature dimension of the running data, and it is used to receive the input running data. There are multiple hidden layers, each consisting of multiple neurons connected by weights. The neurons in the hidden layer receive the output of the previous layer as input, perform linear combination and nonlinear transformation, and then pass the result to the next layer. The output layer is used to output the first feature vector of a set dimension. The running data is input into the support vector machine algorithm to extract features from the running data and obtain a second feature vector; the second feature vector is fused with the first feature vector output by the feature processing module to obtain fused feature data.
[0010] As a further improvement of the present invention, the fault identification module adopts the support vector machine algorithm, random forest algorithm or K-nearest neighbor algorithm.
[0011] As a further improvement of the present invention, the fault identification module is a support vector machine algorithm; the fault identification module specifically includes: The fused feature data is input into the fault identification module. Based on the relationship between the fused features and the operating status of each electronic device, fault classification is performed, and regression prediction is conducted to obtain the fault prediction results for each electronic device. The categories include photovoltaic module faults, inverter faults, cable faults, and combiner box faults. During the training process, the fault identification module learns the relationship between current or voltage characteristics and the operating status of each electronic device based on the hyperplane parameters and kernel function parameters set in the vector machine algorithm.
[0012] As a further improvement of the present invention, the fault prediction results are classified for early warning. When the fault prediction results reach a preset alarm threshold, an automatic alarm mechanism is triggered, specifically including: The warning classification categories include mild warning, moderate warning, and severe warning; The term "mild warning" refers to a notification issued when the prediction results show that the probability of a device malfunction is less than 20% and the malfunction will not affect normal operation. The term "moderate warning" refers to a notification and alert issued when the prediction results indicate that the probability of equipment failure is 20% to 60%, and the equipment failure will affect operation. The severe warning refers to a situation where the prediction results show that the probability of equipment failure is greater than 60%. The failure will not only affect operation but also cause downtime. In this case, a severe warning will be issued and power will be cut off.
[0013] Secondly, the present invention also provides a photovoltaic power plant fault prediction system, comprising: The data acquisition module is used to acquire operating data during the current operation of the photovoltaic electronic equipment. The fault prediction module is used to input the operating data of each electronic device into the trained fault prediction model, and use the fault prediction model to perform fault analysis on the input feature parameters to obtain the fault prediction result. The fault early warning module is used to classify the fault prediction results into warning categories. When the fault prediction results reach the preset alarm threshold, an automatic alarm mechanism is triggered. The fault prediction model includes a feature processing module and a fault identification module. The feature processing module is used to process the input features to obtain feature vectors. The fault identification module is used to perform feature fusion based on the feature vectors and to identify faults based on the fused feature vectors to obtain fault prediction results.
[0014] Thirdly, the present invention also provides a computer-readable storage medium for storing one or more programs, the one or more programs including instructions that, when executed by a computing device, cause the computing device to perform the above-described photovoltaic power plant fault prediction method.
[0015] Fourthly, the present invention also provides a computing device, comprising: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include steps for performing the above-described photovoltaic power plant fault prediction method.
[0016] The beneficial effects of this invention are as follows: The fault prediction method for photovoltaic power plants provided by this invention acquires various operational data, including current, voltage, power generation, environmental data, and inverter operating status, and analyzes them using a fault prediction model. This allows potential problems to be detected before a fault occurs or in its early stages. The feature processing module and fault identification module in the fault prediction model work together. The feature processing module converts various operational data into feature vectors, and the fault identification module then performs feature fusion and fault identification. This mechanism can more accurately locate the fault location (e.g., which photovoltaic module or part of the inverter is malfunctioning) and identify the fault type (e.g., short circuit fault, overheating fault, etc.), greatly improving the accuracy of fault diagnosis compared to traditional methods based on single parameters or simple threshold judgments. The setting of early warning classification and alarm thresholds allows maintenance personnel to rationally allocate resources according to the severity and urgency of the fault. For mild early warnings, routine inspections can be arranged; for moderate early warnings, more manpower and resources can be allocated for detailed inspections; for severe early warnings, emergency repair procedures can be initiated immediately. This avoids over-repair or waste of maintenance resources and improves maintenance efficiency. Photovoltaic module failures directly affect power generation. Timely repair of faulty modules can ensure that the photovoltaic system operates in optimal condition, improve power generation efficiency, thereby increasing power generation and improving the economic benefits of photovoltaic power generation projects. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a flowchart of the photovoltaic power plant fault prediction method in an embodiment of the present invention; Figure 2 This is a schematic diagram of the photovoltaic power plant fault prediction system in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of the computer device in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of the terminal device in an embodiment of the present invention. Detailed Implementation
[0019] To make the objectives and technical solutions of this invention clearer and easier to understand, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. The specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.
[0020] The concept of this invention is that, This system acquires operational data from photovoltaic (PV) electronic devices during a set time period. This data includes at least operating current or voltage data, PV module power generation, environmental data, and inverter operating status data. By acquiring this comprehensive operational data, including current, voltage, power generation, environmental data, and inverter operating status, and analyzing it using a fault prediction model, potential problems can be identified before or in their early stages. For example, in the initial stages of a slight decrease in PV module power generation, it may be due to the formation of hidden cracks or other potential faults. The model can detect this trend in a timely manner and issue an early warning, allowing maintenance personnel to intervene and perform inspections and maintenance in advance, preventing further deterioration of the fault.
[0021] The operating data of each electronic device is input into the trained fault prediction model. The fault prediction model is then used to perform fault analysis on the input feature parameters to obtain the fault prediction results. The fault prediction results are classified into early warning categories. When the fault prediction results reach the preset alarm threshold, the automatic alarm mechanism is triggered. The setting of early warning categories and alarm thresholds allows maintenance personnel to allocate resources rationally based on the severity and urgency of the fault. For minor warnings, routine inspections can be scheduled; for moderate warnings, more manpower and resources can be allocated for detailed checks; and for severe warnings, emergency repair procedures can be initiated immediately. This avoids over-repair or waste of maintenance resources, improving maintenance efficiency.
[0022] The prediction model includes a feature processing module and a fault identification module. The feature processing module processes the input features to obtain feature vectors; the fault identification module performs feature fusion based on the feature vectors, and identifies faults based on the fused feature vectors to obtain fault prediction results.
[0023] The feature processing module and fault identification module in the fault prediction model work together. The feature processing module transforms various operational data into feature vectors, and the fault identification module then performs feature fusion and fault identification. This mechanism can more accurately locate the fault location (such as which photovoltaic module or part of the inverter is malfunctioning) and identify the fault type (such as short circuit fault, overheating fault, etc.), greatly improving the accuracy of fault diagnosis compared to traditional methods based on a single parameter or simple threshold judgment.
[0024] Continuous monitoring of operational data and fault prediction reduces the probability of faults and downtime. For example, timely detection and repair of potential faults in photovoltaic modules ensures continuous and stable power generation from the photovoltaic system, improving the overall system reliability. This is crucial for the stable power supply of photovoltaic power plants, especially in scenarios with high requirements for power supply continuity, such as industrial power or off-grid photovoltaic power generation systems.
[0025] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings and specific embodiments. The described embodiments are only some embodiments of the present invention, and not all embodiments.
[0026] Example 1 like Figure 1 As shown, the present invention provides a method for predicting faults in photovoltaic power plants, which specifically includes the following implementation steps.
[0027] To ensure efficient operation of photovoltaic electronic equipment within a preset time period, and to obtain the operating data of the photovoltaic electronic equipment during the current set time period, we need to collect the following key data, including at least: operating current or voltage data, photovoltaic module power generation, environmental data, and inverter operating status data.
[0028] Data on operating current or voltage, power generation from photovoltaic modules, environmental parameters (including but not limited to light intensity, temperature, and humidity), and inverter operating status (e.g., operating frequency, output power, and fault codes) should be collected from the equipment's sensors and monitoring systems.
[0029] Data collected from the equipment's sensors and monitoring systems includes operating current or voltage data, photovoltaic module power generation, environmental data (such as light intensity, temperature, humidity, etc.), and inverter operating status data (such as operating frequency, output power, fault codes, etc.).
[0030] After acquiring the operational data during the current operation of the photovoltaic electronic device, the process further includes preprocessing the operational data based on historical datasets. Preprocessing includes denoising the acquired operational data and normalizing the denoised operational data. The normalization process includes:
[0031] In the formula, X represents the currently running data. The minimum value among the running data in the historical dataset. This represents the maximum value in the historical dataset.
[0032] The operating data of each electronic device is input into the trained fault prediction model. The fault prediction model is then used to analyze the input feature parameters to obtain the fault prediction results.
[0033] The fault prediction model includes a feature processing module and a fault identification module. The feature processing module processes the input features to obtain feature vectors; the fault identification module performs feature fusion based on the feature vectors and identifies faults based on the fused feature vectors to obtain the fault prediction result.
[0034] The feature processing module employs convolutional neural networks, multilayer perceptrons, support vector machines, and recurrent neural networks.
[0035] The feature processing module employs multilayer perceptron and support vector machine algorithms; The multilayer perceptron algorithm includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is the same as the feature dimension of the running data, and it is used to receive the input running data. There are multiple hidden layers, each consisting of multiple neurons connected by weights. The neurons in the hidden layer receive the output of the previous layer as input, perform linear combination and nonlinear transformation, and then pass the result to the next layer. The output layer is used to output the first feature vector of a set dimension.
[0036] In the multilayer perceptron algorithm, the input layer plays a crucial role in data reception throughout the feature processing flow. The number of neurons in the input layer is the same as the feature dimensions of the operational data; this design ensures the complete reception of the input operational data. For example, when the operational data includes operating current, voltage data, photovoltaic module power generation, environmental data (such as multiple dimensions of data like light intensity, temperature, and humidity), and inverter operating status data (such as multiple parameters like operating frequency and output power), the input layer will have corresponding neurons to match these data, thereby accurately acquiring this information.
[0037] There are multiple hidden layers, each composed of multiple neurons. Neurons are connected by weights, and this connection method determines the rules for data transfer and transformation within the hidden layers. In each hidden layer, a neuron receives the output of the previous layer as input and first performs a linear combination operation. For example, for the i-th neuron in the j-th hidden layer, it will linearly combine the outputs x1, x2, ..., xn (n is the number of neurons in the previous layer) according to their weights to obtain... Then, a nonlinear transformation is applied to the result of the linear combination. Commonly used nonlinear transformation functions include the Sigmoid function and the ReLU function. For example, if the ReLU function is used, the result after the nonlinear transformation is y=max(0,z). After this processing, the neuron passes the result to the next layer of neurons. Through the processing of multiple hidden layers, the multilayer perceptron can perform deep feature extraction and transformation on the input data, uncovering the complex relationships within the data.
[0038] The output layer outputs a first feature vector with a defined dimension. This defined dimension is determined based on the requirements of the subsequent fault identification module and the architecture of the entire fault prediction model. The output layer integrates and maps the results processed by the hidden layers to obtain a first feature vector with a specific dimension. This first feature vector contains feature information extracted from the running data and processed by the multilayer perceptron algorithm; this information will serve as an important component of subsequent feature fusion.
[0039] The running data is input into the support vector machine algorithm, and the feature processing module is used to extract features from the running data to obtain a second feature vector. The second feature vector is fused with the first feature vector output by the feature processing module to obtain fused feature data.
[0040] The operational data is input into the Support Vector Machine (SVM) algorithm, which extracts features from the data using its unique mathematical principles. The SVM algorithm searches for the optimal classification hyperplane in a high-dimensional space, uncovering key features from the operational data in the process. For example, for operational data of photovoltaic electronic devices, it can identify which data features are significant in distinguishing between normal and fault states. In this way, the SVM algorithm obtains a second feature vector, which contains the results of feature extraction from the operational data from the SVM's perspective.
[0041] After obtaining the first feature vector output by the multilayer perceptron algorithm and the second feature vector obtained by the support vector machine algorithm, the two are fused. Feature fusion involves integrating two vectors from different sources, both related to the features of the running data, to obtain fused feature data. This fusion method fully leverages the advantages of the multilayer perceptron algorithm in uncovering complex relationships within the data and the feature extraction characteristics of the support vector machine algorithm, resulting in fused feature data containing more comprehensive and valuable information, thus providing strong support for accurate fault identification.
[0042] The fault identification module uses the support vector machine algorithm, random forest algorithm, or K-nearest neighbor algorithm.
[0043] In this embodiment, the fault identification module uses a support vector machine algorithm. The fault identification module specifically includes: The fused feature data is input into the fault identification module. Based on the relationship between the fused features and the operating status of each electronic device, fault classification is performed, and regression prediction is conducted to obtain the fault prediction results for each electronic device. The categories include photovoltaic module faults, inverter faults, cable faults, and combiner box faults, etc. During the training process, the fault identification module learns the relationship between current or voltage characteristics and the operating status of each electronic device based on the hyperplane parameters and kernel function parameters set in the vector machine algorithm.
[0044] Fault prediction results are categorized for early warning. When the fault prediction results reach a preset alarm threshold, an automatic alarm mechanism is triggered. Specifically, this includes: The warning categories include mild warning, moderate warning, and severe warning; A mild warning refers to a notification issued when the prediction result shows that the probability of a device failure is less than 20%, and the failure will not affect normal operation. For example, if a neural network and support vector machine fusion model predicts that the probability of a slight power degradation failure in photovoltaic modules in the future is 15%, the equipment may be in an early stage of performance fluctuation, possibly due to temporary effects from environmental factors (such as slight temperature changes or brief shading). A moderate warning refers to a notification and alert issued when the prediction shows a 20% to 60% probability of equipment failure, which would affect operation. For example, if the predicted probability of an inverter experiencing an overcurrent fault in the near future is 40%, it may be due to factors such as aging of some circuit components or slight load fluctuations. Although no fault has occurred yet, the situation is relatively serious and requires further inspection and evaluation.
[0045] A severe warning refers to a situation where the prediction result indicates that the probability of equipment failure is greater than 60%. This failure would not only affect operation but also cause downtime. In such cases, a severe warning notification and power outage are issued. For example, if the predicted probability of hot spot failure in photovoltaic cells reaches 80%, it may be due to severe localized shading or internal defects in the cells. Immediate action is necessary in such situations; otherwise, permanent damage to the cells may occur, or even fires and other safety issues.
[0046] The information in the early warning notification includes: Equipment identification: Clearly indicate which photovoltaic electronic device is experiencing a potential malfunction, such as the specific photovoltaic module number, inverter identification, etc.
[0047] Fault type prediction: Provide a detailed description of the predicted fault type, such as "photovoltaic module hot spot fault" or "inverter overvoltage fault".
[0048] Fault probability or severity: Inform maintenance personnel of the probability of a fault occurring or the severity of the warning, such as "the probability of a fault occurring is 40% (moderate warning)".
[0049] Relevant parameter values: List the current values of operating parameters and alarm thresholds related to the fault, such as "Current output power is 800W, rated power is 1000W, power drop has reached 20% (triggered severe warning)".
[0050] Automatic alarm mechanisms can be implemented in various ways, such as displaying detailed fault information (including device name, fault type, and warning level) in a pop-up alarm window within the monitoring system, sending SMS or emails to maintenance personnel, or activating audible and visual alarms. Upon receiving an alarm notification, maintenance personnel should promptly review the alarm information to understand the general situation of the fault, including the device that experienced the fault, the fault type, and the warning level.
[0051] This embodiment also dynamically adjusts the alarm threshold based on the equipment's historical operating data and environmental factors. If the equipment's normal operating parameters differ significantly between summer and winter, the alarm threshold can be adjusted according to seasonal factors. For example, in the high-temperature environment of summer, the normal temperature range of photovoltaic modules is higher than in winter, so the temperature alarm threshold can be appropriately increased in summer; while in seasons with weaker sunlight, the power output alarm threshold can be adjusted according to the average power level of that season.
[0052] As equipment operates for longer periods, its performance gradually declines, necessitating adjustments to alarm thresholds. For photovoltaic electronic equipment that has been in operation for many years, performance degradation is normal, and this factor should be considered when setting alarm thresholds. For example, for photovoltaic modules that have been operating for more than 5 years, the power degradation alarm threshold can be relaxed from the initial 20% to 25% to accommodate the normal aging process of the equipment.
[0053] Example 2 like Figure 2 As shown, another embodiment of the present invention provides a photovoltaic power plant fault prediction system for implementing the photovoltaic power plant fault prediction method in Embodiment 1. The system includes: The data acquisition module is used to acquire operating data during the current operation of the photovoltaic electronic equipment. The fault prediction module is used to input the operating data of each electronic device into the trained fault prediction model, and use the fault prediction model to perform fault analysis on the input feature parameters to obtain the fault prediction result. The fault early warning module is used to classify the fault prediction results into warning categories. When the fault prediction results reach the preset alarm threshold, an automatic alarm mechanism is triggered. The fault prediction model includes a feature processing module and a fault identification module. The feature processing module is used to process the input features to obtain feature vectors. The fault identification module is used to perform feature fusion based on the feature vectors and to identify faults based on the fused feature vectors to obtain fault prediction results.
[0054] Example 3 In another embodiment of the present invention, a terminal device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to achieve a corresponding method flow or corresponding function. The processor described in this embodiment of the present invention can be used in the operation of a photovoltaic power station fault prediction method, including: Acquire the operating data of the photovoltaic electronic equipment during the current set time period. The operating data includes at least the operating current or voltage data, the power generation of the photovoltaic module, environmental data, and the inverter operating status data. The operating data of each electronic device is input into the trained fault prediction model. The fault prediction model is then used to perform fault analysis on the input feature parameters to obtain the fault prediction results. The fault prediction results are classified for early warning. When the fault prediction results reach the preset alarm threshold, an automatic alarm mechanism is triggered. The fault prediction model includes a feature processing module and a fault identification module. The feature processing module is used to process the input features to obtain feature vectors. The fault identification module is used to perform feature fusion based on the feature vectors and to identify faults based on the fused feature vectors to obtain fault prediction results.
[0055] Example 4 In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and extended storage media supported by the terminal device; it can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor, which can be one or more computer programs (including program code). It should be noted that more specific examples (a non-exhaustive list) of the computer-readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
[0056] Computer-readable storage media also include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium can also be any readable medium other than a readable storage medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium can be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.
[0057] Program code for performing the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages such as Java and C++, and conventional procedural programming languages such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).
[0058] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the photovoltaic power plant fault prediction method in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor in the following steps: Acquire the operating data of the photovoltaic electronic equipment during the current set time period. The operating data includes at least the operating current or voltage data, the power generation of the photovoltaic module, environmental data, and the inverter operating status data. The operating data of each electronic device is input into the trained fault prediction model. The fault prediction model is then used to perform fault analysis on the input feature parameters to obtain the fault prediction results. The fault prediction results are classified for early warning. When the fault prediction results reach the preset alarm threshold, an automatic alarm mechanism is triggered. The fault prediction model includes a feature processing module and a fault identification module. The feature processing module is used to process the input features to obtain feature vectors. The fault identification module is used to perform feature fusion based on the feature vectors and to identify faults based on the fused feature vectors to obtain fault prediction results.
[0059] Figure 3 This is a schematic diagram of a computer device provided according to an embodiment of the present invention.
[0060] Please see Figure 3 The terminal device is a computer device. In this embodiment, the computer device 60 includes a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When executed by the processor 61, the computer program 63 implements the fluid composition calculation method in the reservoir stimulation wellbore of this embodiment. To avoid repetition, these details are not elaborated here. Alternatively, when executed by the processor 61, the computer program 63 implements the functions of each model / unit in the composition calculation system of this embodiment. To avoid repetition, these details are not elaborated here.
[0061] Computer device 60 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. Computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will understand that... Figure 3 This is merely an example of computer device 60 and does not constitute a limitation on computer device 60. It may include more or fewer components than shown, or combine certain components, or different components. For example, computer device may also include input / output devices, network access devices, buses, etc.
[0062] The processor 61 may be a central processing unit (CPU), or other general-purpose processors, CPUs, graphics processing units (GPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, quantum computing-based data processing logic units, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.
[0063] The memory 62 can be an internal storage unit of the computer device 60, such as a hard disk or RAM of the computer device 60. The memory 62 can also be an external storage device of the computer device 60, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc., equipped on the computer device 60.
[0064] Furthermore, the memory 62 may include both internal storage units of the computer device 60 and external storage devices. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 can also be used to temporarily store data that has been output or will be output.
[0065] Any references to memory, databases, or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.
[0066] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.
[0067] Figure 4 This is a block diagram of an electronic device according to an embodiment of the present invention.
[0068] Please see Figure 4 The terminal device 600 is an electronic device, which takes the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including storage unit 620 and processing unit 610), a display unit 640, etc.
[0069] The storage unit stores program code, which can be executed by the processing unit 610 to perform the steps described in the method section of this specification according to various exemplary embodiments of the present invention. For example, the processing unit 610 can perform actions such as... Figure 1 The steps are shown in the figure.
[0070] Storage unit 620 may include a readable medium in the form of a volatile storage unit, such as random access memory (RAM) 6201 and / or cache memory 6202, and may further include a read-only memory (ROM) 6203.
[0071] Storage unit 620 may also include a program / utility 6204 having a set (at least one) program module 6205, such program module 6205 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.
[0072] Bus 630 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the multiple bus structures.
[0073] Electronic device 600 can also communicate with one or more external devices 700 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable a user to interact with electronic device 600, and / or with any device that enables electronic device 600 to communicate with one or more other computing devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 650. Furthermore, electronic device 600 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 660. Network adapter 660 can communicate with other modules of electronic device 600 via bus 630. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.
[0074] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
Claims
1. A method for predicting faults in a photovoltaic power plant, characterized in that, include: The system acquires operational data of photovoltaic electronic devices during the current set time period. The operational data includes at least operational current or voltage data, power generation of photovoltaic modules, environmental data, and inverter operating status data. The operating data of each electronic device is input into the trained fault prediction model. The fault prediction model is then used to perform fault analysis on the input feature parameters to obtain the fault prediction results. The fault prediction results are classified for early warning. When the fault prediction results reach the preset alarm threshold, an automatic alarm mechanism is triggered. The fault prediction model includes a feature processing module and a fault identification module. The feature processing module is used to perform feature processing on the input to obtain a feature vector. The fault identification module is used to perform feature fusion based on feature vectors, and to identify faults based on the fused feature vectors to obtain fault prediction results.
2. The photovoltaic power plant fault prediction method according to claim 1, characterized in that, After acquiring the operating data during the current operation of the photovoltaic electronic device, the method further includes preprocessing the operating data based on historical datasets. The preprocessing includes denoising the acquired operating data and normalizing the denoised operating data. The normalization process includes: In the formula, X represents the currently running data. The minimum value among the running data in the historical dataset. This represents the maximum value in the historical dataset.
3. The photovoltaic power plant fault prediction method according to claim 1, characterized in that, The feature processing module employs convolutional neural networks, multilayer perceptrons, support vector machines, or recurrent neural networks.
4. The photovoltaic power plant fault prediction method according to claim 3, characterized in that, The feature processing module employs a multilayer perceptron algorithm and a support vector machine algorithm. The multilayer perceptron algorithm includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is the same as the feature dimension of the running data, and it is used to receive the input running data. There are multiple hidden layers, each consisting of multiple neurons connected by weights. The neurons in the hidden layer receive the output of the previous layer as input, perform linear combination and nonlinear transformation, and then pass the result to the next layer. The output layer is used to output the first feature vector of a set dimension. The running data is input into the support vector machine algorithm to extract features from the running data and obtain a second feature vector; the second feature vector is fused with the first feature vector output by the feature processing module to obtain fused feature data.
5. The photovoltaic power plant fault prediction method according to claim 1, characterized in that, The fault identification module employs a support vector machine algorithm, a random forest algorithm, or a K-nearest neighbor algorithm.
6. The photovoltaic power plant fault prediction method according to claim 5, characterized in that, The fault identification module uses a support vector machine algorithm; the fault identification module specifically includes: The fused feature data is input into the fault identification module. Based on the relationship between the fused features and the operating status of each electronic device, fault classification is performed, and regression prediction is conducted to obtain the fault prediction results for each electronic device. The categories include photovoltaic module faults, inverter faults, cable faults, and combiner box faults. During the training process, the fault identification module learns the relationship between current or voltage characteristics and the operating status of each electronic device based on the hyperplane parameters and kernel function parameters set in the vector machine algorithm.
7. The photovoltaic power plant fault prediction method according to claim 1, characterized in that, The fault prediction results are classified for early warning. When the fault prediction results reach a preset alarm threshold, an automatic alarm mechanism is triggered, specifically including: The warning classification categories include mild warning, moderate warning, and severe warning; The term "mild warning" refers to a notification issued when the prediction results show that the probability of a device malfunction is less than 20% and the malfunction will not affect normal operation. The term "moderate warning" refers to a notification and alert issued when the prediction results indicate that the probability of equipment failure is 20% to 60%, and the equipment failure will affect operation. The severe warning refers to a situation where the prediction results show that the probability of equipment failure is greater than 60%. The failure will not only affect operation but also cause downtime. In this case, a severe warning will be issued and power will be cut off.
8. A photovoltaic power plant fault prediction system, characterized in that, include: The data acquisition module is used to acquire operating data during the current operation of the photovoltaic electronic equipment. The fault prediction module is used to input the operating data of each electronic device into the trained fault prediction model, and use the fault prediction model to perform fault analysis on the input feature parameters to obtain the fault prediction result. The fault early warning module is used to classify the fault prediction results into warning categories. When the fault prediction results reach the preset alarm threshold, an automatic alarm mechanism is triggered. The fault prediction model includes a feature processing module and a fault identification module. The feature processing module is used to perform feature processing on the input to obtain a feature vector. The fault identification module is used to perform feature fusion based on feature vectors, and to identify faults based on the fused feature vectors to obtain fault prediction results.
9. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform the photovoltaic power plant fault prediction method according to any one of claims 1 to 7.
10. A computing device, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including steps for performing the photovoltaic power plant fault prediction method of any one of claims 1 to 7.