Fault detection method and device for secondary protection screen cabinet, terminal equipment and storage medium

Abstract

The invention discloses a fault detection method and device for a secondary protection screen cabinet, terminal equipment and a storage medium, and the method comprises the steps: inputting time sequence data and working condition data of the secondary protection screen cabinet into a fault detection model, and enabling the fault detection model to extract electrical features in the time sequence data and working condition features in the working condition data; according to the method, the characteristics of the secondary protection screen cabinet are extracted, and the characteristics are fused to generate multi-dimensional fusion characteristics, so that the fault type and the fault early warning degree of the secondary protection screen cabinet can be automatically and quickly obtained based on the fusion characteristics, and the model can comprehensively consider the influence of various factors on equipment faults through a characteristic fusion method. Therefore, the fault identification precision is improved. If the fault type and the fault early warning degree of the secondary protection screen cabinet are detected, the fault type and the fault early warning degree of the secondary protection screen cabinet can be rapidly and accurately obtained through model identification and processing, corresponding fault early warning information is automatically generated, potential fault problems can be rapidly identified, and the fault maintenance efficiency of the secondary protection screen cabinet is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault detection of power systems, and in particular to a fault detection method, device, terminal equipment and storage medium for a secondary protection panel cabinet. Background Art

[0002] The secondary protection panel cabinet is one of the key devices used to protect and control electrical equipment in the power system. It contains a variety of protection relays, control switches, indicating instruments, etc., which are used to monitor the voltage, current and other parameters in the power system. When a fault occurs in the power system, it can quickly cut off the fault source and protect the safe and stable operation of the power grid and equipment.

[0003] The stability and reliability of the secondary protection cabinet as a protective barrier for the power system are crucial to the safe operation of the power grid. Fault detection can timely discover and eliminate potential problems, ensure that the protection cabinet can operate correctly and quickly when needed, and prevent the further expansion of the fault.

[0004] However, traditional fault identification and maintenance often rely on manual inspections and empirical judgment, which is inefficient and prone to errors. Existing technologies are unable to quickly identify potential problems, resulting in low maintenance efficiency. Summary of the invention

[0005] The embodiments of the present invention provide a fault detection method, device, terminal equipment and storage medium for a secondary protection panel cabinet. Through model identification and processing, the fault type and fault warning degree of the secondary protection panel cabinet can be quickly and accurately obtained, which can effectively solve the problem of low maintenance efficiency caused by relying on manual inspection and experience judgment in the prior art.

[0006] An embodiment of the present invention provides a secondary protection screen cabinet fault detection method, comprising:

[0007] Obtaining the time series data and operating condition data corresponding to the secondary protection panel cabinet; wherein the time series data includes: voltage data and current data; the operating condition data includes: environmental index values ​​and operating parameters;

[0008] Input the time series data and the operating condition data into a fault detection model, so that the fault detection model extracts the electrical features in the time series data and the operating condition features in the operating condition data, and fuses the electrical features with the operating condition features to generate multi-dimensional fusion features, and then generates the fault type and fault warning degree of the secondary protection panel cabinet according to the fusion features;

[0009] Generate fault warning information corresponding to the secondary protection panel cabinet according to the fault type and fault warning degree;

[0010] The generation of the fault detection model includes:

[0011] Each time series sample and each working condition sample of the secondary protection cabinet are used as corresponding training samples;

[0012] Taking each training sample, the actual fault type corresponding to each training sample and the actual fault warning degree corresponding to each training sample as input, and the predicted fault type of each training sample and the predicted fault warning degree of each training sample as output, the fault detection model to be trained is iteratively trained until the model converges to generate a fault detection model.

[0013] Preferably, the fault detection model comprises: a feature extraction layer, a feature fusion layer and a classification layer;

[0014] The feature extraction layer is used to extract electrical features based on voltage data and current data in the time series data, and to extract operating condition features in the operating condition data;

[0015] The feature fusion layer is used to perform nonlinear transformation on the electrical features and the operating condition features to generate fusion features with multiple dimensions;

[0016] The classification layer is used to generate the fault type and fault warning degree of the secondary protection screen cabinet according to the fusion characteristics.

[0017] Preferably, the working condition data further includes: communication parameters and attitude parameters; the communication parameters include: signal strength and data transmission rate; the attitude parameters include: device tilt angle and vibration amplitude;

[0018] The feature extraction layer is used to extract electrical features based on voltage data and current data in the time series data, and to extract operating condition features in the operating condition data, including:

[0019] The feature extraction layer performs temporal modeling on the time series data to capture long-term dependencies in the sequence;

[0020] Through the long-term dependency in the sequence, a first hidden state sequence and a second hidden state sequence corresponding to the voltage data and the current data are output respectively; wherein the first hidden state sequence is used to represent the vector representation of the voltage in the time dimension; and the second hidden state sequence is used to represent the vector representation of the current in the time dimension;

[0021] generating an electrical signature according to the first hidden state sequence and the second hidden state sequence;

[0022] Based on the environmental index values, operating parameters, communication parameters, posture parameters, equipment tilt angles, and vibration amplitudes, operating condition characteristics for characterizing operating state changes of the secondary protection panel cabinet are extracted.

[0023] Preferably, the feature fusion layer is used to perform nonlinear transformation on the electrical features and the operating condition features to generate multi-dimensional fusion features, including:

[0024] The feature fusion layer is used to assign different weights to the electrical features and the operating condition features, and then linearly combine the electrical features with the assigned weights and the operating condition features with the assigned weights;

[0025] The activation function is used to perform nonlinear transformation on the features of the linear combination to generate multi-dimensional fusion features.

[0026] Preferably, the classification layer is used to generate the fault type and fault warning degree of the secondary protection screen cabinet according to the fusion feature, including:

[0027] The classification layer is used to calculate the similarity between the fusion feature and the weight of each fault type, convert the obtained similarity into probability, and output probability values ​​of several fault types;

[0028] The probability values ​​of various fault types are compared, the fault type with the highest probability value is taken as the fault type of the secondary protection panel cabinet, and the fault warning degree corresponding to the fault type with the highest probability value is output.

[0029] Preferably, the feature fusion layer is used to assign different weights to the electrical features and the operating condition features, including:

[0030] The attention mechanism layer in the feature fusion layer calculates initial weights corresponding to the electrical features and the operating condition features, respectively, based on the input electrical features and the operating condition features; wherein the initial weights are used to characterize the contribution of the features to the fault type;

[0031] The electrical characteristics and the operating condition characteristics are weighted and fused with the corresponding initial weights respectively to generate the weights corresponding to the electrical characteristics and the weights corresponding to the operating condition characteristics.

[0032] Preferably, during each iterative training, a training sample is input into the fault detection model, so that the fault detection model generates a predicted fault type and a predicted fault warning degree corresponding to the training sample according to the electrical characteristics and operating condition characteristics in the training sample;

[0033] The predicted fault type and the predicted fault warning degree are compared with the actual fault type and the actual fault warning degree of the training sample, and then the network parameters of the fault detection model are adjusted according to the comparison results.

[0034] Based on the above method embodiments, the present invention provides corresponding device embodiments.

[0035] An embodiment of the present invention provides a fault detection device for a secondary protection screen cabinet, comprising: a data acquisition module, a fault type determination module and an early warning module;

[0036] The data acquisition module is used to acquire the time series data and operating condition data corresponding to the secondary protection panel cabinet; wherein the time series data includes: voltage data and current data; the operating condition data includes: environmental index values ​​and operating parameters;

[0037] The fault type determination module is used to input the time series data and the operating condition data into a fault detection model, so that the fault detection model extracts the electrical features in the time series data and the operating condition features in the operating condition data, and fuses the electrical features with the operating condition features to generate multi-dimensional fusion features, and then generates the fault type and fault warning degree of the secondary protection panel cabinet according to the fusion features;

[0038] The warning module is used to generate fault warning information corresponding to the secondary protection screen cabinet according to the fault type and fault warning degree;

[0039] The generation of the fault detection model includes:

[0040] Each time series sample and each working condition sample of the secondary protection cabinet are used as corresponding training samples;

[0041] Taking each training sample, the actual fault type corresponding to each training sample and the actual fault warning degree corresponding to each training sample as input, and the predicted fault type of each training sample and the predicted fault warning degree of each training sample as output, the fault detection model to be trained is iteratively trained until the model converges to generate a fault detection model.

[0042] Based on the above method embodiments, the present invention provides corresponding terminal device embodiments.

[0043] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the fault detection method of a secondary protection screen cabinet described in the above-mentioned embodiment of the invention.

[0044] Based on the above method embodiments, the present invention provides corresponding storage medium item embodiments.

[0045] Another embodiment of the present invention provides a storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute a secondary protection screen cabinet fault detection method described in the above-mentioned invention embodiment.

[0046] The following beneficial effects are achieved by implementing the present invention:

[0047] The embodiment of the present invention provides a fault detection method, device, terminal device and storage medium for a secondary protection screen cabinet. The present invention obtains the time series data of the secondary protection screen cabinet, such as voltage and current data, and working condition data, such as environmental index values ​​and operating parameters, so that after these time series data and working condition data are input into the fault detection model, the fault detection model can extract the electrical characteristics in the time series data and the working condition characteristics in the working condition data, and fuse these characteristics to generate multi-dimensional fusion characteristics, so that based on the fusion characteristics, the fault type and fault warning degree of the secondary protection screen cabinet can be automatically and quickly obtained, and the above-mentioned feature fusion method enables the model to comprehensively consider the influence of multiple factors on equipment failure, thereby improving the accuracy of fault identification. Compared with the prior art, the present invention can be based on the time series data and working condition data corresponding to the secondary protection screen cabinet, through model identification and processing, can quickly and accurately obtain the fault type and fault warning degree of the secondary protection screen cabinet, and then automatically generate corresponding fault warning information, so the present invention can quickly identify potential fault problems and improve the fault inspection and repair efficiency of the secondary protection screen cabinet. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flow chart of a fault detection method for a secondary protection screen cabinet provided in one embodiment of the present invention.

[0049] Figure 2 It is a structural schematic diagram of a fault detection device for a secondary protection screen cabinet provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] like Figure 1 FIG. 1 is a flow chart of a fault detection method for a secondary protection screen cabinet provided by an embodiment of the present invention. The fault detection method for a secondary protection screen cabinet includes:

[0052] Step S1: Obtaining the time series data and operating condition data corresponding to the secondary protection panel cabinet; wherein the time series data includes: voltage data and current data; the operating condition data includes: environmental index values ​​and operating parameters;

[0053] Step S2: inputting the time series data and the operating condition data into a fault detection model, so that the fault detection model extracts the electrical features in the time series data and the operating condition features in the operating condition data, and fuses the electrical features with the operating condition features to generate multi-dimensional fusion features, and then generates the fault type and fault warning degree of the secondary protection panel cabinet according to the fusion features;

[0054] The generation of the fault detection model includes:

[0055] Each time series sample and each working condition sample of the secondary protection cabinet are used as corresponding training samples;

[0056] Taking each training sample, the actual fault type corresponding to each training sample and the actual fault warning degree corresponding to each training sample as input, and taking the predicted fault type of each training sample and the predicted fault warning degree of each training sample as output, the fault detection model to be trained is iteratively trained until the model converges to generate a fault detection model.

[0057] Step S3: Generate fault warning information corresponding to the secondary protection panel cabinet according to the fault type and fault warning degree.

[0058] For step S1, in a preferred embodiment, the working condition data of the present invention further includes: communication parameters and attitude parameters; the communication parameters include: signal strength and data transmission rate; the attitude parameters include: device tilt angle and vibration amplitude.

[0059] The present invention can collect multi-dimensional data of the secondary protection panel cabinet during the fault detection process, such as voltage timing data, current timing data, environmental indicator values ​​(such as temperature, humidity and other environmental indicators), communication parameters (such as signal strength, data transmission rate and other communication status indicators) and posture parameters (such as equipment tilt angle, vibration amplitude and other physical status indicators).

[0060] Since multi-dimensional data can more comprehensively reflect the operating status and external environmental conditions of the secondary protection panel cabinet, the present invention can enable the model to more accurately identify potential faults or abnormal conditions by collecting these data and using them as inputs to the model.

[0061] For step S2, in a preferred embodiment, the time series data and the operating condition data are input into a fault detection model, and the fault detection model generates the fault type and the fault warning degree of the secondary protection panel cabinet.

[0062] The generation and training process of the fault detection model also includes:

[0063] In each iterative training, a training sample is input into the fault detection model, so that the fault detection model generates the predicted fault type and predicted fault warning degree corresponding to the training sample according to the electrical characteristics and working condition characteristics in the training sample;

[0064] The predicted fault type and the predicted fault warning degree are compared with the actual fault type and the actual fault warning degree of the training sample, and then the network parameters of the fault detection model are adjusted according to the comparison results.

[0065] Specifically, the training process of the fault detection model of the present invention is:

[0066] Collect a large amount of time series data (such as voltage and current time series data) and working condition data (such as environmental index values, communication parameters, posture parameters, etc.) of the secondary protection cabinet and use them as samples. Preprocess the data in the samples, including data cleaning, normalization, feature extraction and other steps to ensure the accuracy and consistency of the data.

[0067] The neural network structure is used as the basic framework of the fault detection model to initialize the network parameters of the model, and these parameters can be continuously adjusted during the training process to optimize the model performance.

[0068] The training samples (including time series data and operating condition data) are input into the model one by one, so that the model performs forward propagation according to the input characteristics (electrical characteristics and operating condition characteristics), and finally generates the predicted fault type and predicted fault warning degree.

[0069] The prediction results are then compared with the actual fault types and actual fault warning levels of the training samples, and the loss function (such as cross entropy loss, mean square error, etc.) is calculated to quantify the prediction error. Then, back propagation is performed based on the gradient of the loss function, and the network parameters of the model are adjusted through optimization algorithms such as gradient descent to reduce the prediction error.

[0070] Repeat the above steps until the performance of the model on the validation set no longer improves significantly, and the model is considered to have converged.

[0071] Therefore, through iterative training, the fault detection model of the present invention can continuously learn from data and optimize its prediction ability, thereby improving the accuracy of fault detection. The comprehensive use of multi-dimensional data enables the fault detection model to more comprehensively understand the operating status and external environmental conditions of the equipment, and thus more accurately identify the fault type and warning level.

[0072] The trained fault detection model can be deployed in the actual system for real-time detection. Therefore, when new time series data and operating condition data are input, the fault detection model can quickly generate prediction results and issue corresponding fault warning information.

[0073] The present invention can detect and handle potential faults in a timely manner through real-time detection, thus preventing the faults from having a greater impact on the system. Based on the prediction results of the model, maintenance personnel can be assisted in formulating a more scientific and reasonable equipment maintenance strategy. For example, preventive maintenance tasks can be arranged according to the level of fault warning to reduce unplanned downtime and maintenance costs, thereby improving the operating efficiency and economic benefits of the entire system.

[0074] In a preferred embodiment, the fault detection model of the embodiment of the present invention includes: a feature extraction layer, a feature fusion layer and a classification layer;

[0075] The feature extraction layer is used to extract electrical features based on voltage data and current data in the time series data, and to extract operating condition features in the operating condition data;

[0076] The feature fusion layer is used to perform nonlinear transformation on the electrical features and the operating condition features to generate fusion features with multiple dimensions;

[0077] The classification layer is used to generate the fault type and fault warning degree of the secondary protection screen cabinet according to the fusion characteristics.

[0078] Specifically, when the feature extraction layer extracts features, due to the working condition data, it also includes: communication parameters and posture parameters; the communication parameters include: signal strength and data transmission rate; the posture parameters include: device tilt angle and vibration amplitude.

[0079] The feature extraction layer can perform temporal modeling on the time series data to capture long-term dependencies in the sequence;

[0080] Through the long-term dependency in the sequence, a first hidden state sequence and a second hidden state sequence corresponding to the voltage data and the current data are output respectively; wherein the first hidden state sequence is used to represent the vector representation of the voltage in the time dimension; and the second hidden state sequence is used to represent the vector representation of the current in the time dimension;

[0081] generating an electrical signature according to the first hidden state sequence and the second hidden state sequence;

[0082] Based on the environmental index values, operating parameters, communication parameters, posture parameters, equipment tilt angles, and vibration amplitudes, operating condition characteristics for characterizing operating state changes of the secondary protection panel cabinet are extracted.

[0083] In a preferred embodiment, the feature extraction layer is an LSTM layer, which can capture long-term dependencies in time series data, and the LSTM layer updates and transmits internal states through its unique gating mechanism (forget gate, input gate, output gate), thereby effectively processing time series data and extracting key features from each data.

[0084] Since the LSTM layer is good at processing time series data, it can capture long-term dependencies in the data. In the fault detection of the secondary protection cabinet, the time series data of voltage and current often contain rich historical information, which is crucial for identifying potential faults. The LSTM layer can effectively retain and utilize this historical information through its internal state update and transmission mechanism, thereby improving the accuracy of fault detection.

[0085] Moreover, the LSTM layer can also process noise and outliers in time series data, so feature extraction through the LSTM layer can enhance the robustness of the model. In practical applications, the time series data and working condition data of the secondary protection cabinet may be affected by various interference factors, such as sensor noise, communication interference, etc. The LSTM layer can suppress the influence of these interference factors to a certain extent through the smooth update mechanism of its internal state, thereby improving the stability and reliability of the model.

[0086] Furthermore, when generating fusion features with multiple dimensions, the feature fusion layer can assign different weights to the electrical features and the operating condition features, and then linearly combine the electrical features with the assigned weights and the operating condition features with the assigned weights;

[0087] Schematically, the attention mechanism layer in the feature fusion layer can calculate the initial weights corresponding to the electrical features and the operating condition features according to the input electrical features and the operating condition features; wherein the initial weights are used to characterize the contribution of the features to the fault type;

[0088] The electrical characteristics and the operating condition characteristics are weighted and fused with the corresponding initial weights respectively to generate the weights corresponding to the electrical characteristics and the weights corresponding to the operating condition characteristics.

[0089] The electrical features with assigned weights and the operating condition features with assigned weights can be further linearly combined, and the features of the linear combination can be nonlinearly transformed through an activation function to generate multi-dimensional fusion features.

[0090] In the feature fusion layer, by assigning different weights to electrical features and operating condition features and performing linear combinations, combined with nonlinear transformations, more expressive and discriminative fusion features can be generated. The embodiment of the present invention not only considers the different contributions of different features to fault types, but also enhances the complexity of features through nonlinear transformations, thereby improving the accuracy and robustness of fault detection.

[0091] Specifically, the attention mechanism layer in the feature fusion layer also plays a key role. Since the attention mechanism can automatically learn and calculate the initial weights corresponding to the electrical features and working condition features, and these weights reflect the importance or contribution of each feature in identifying the fault type, weighted fusion, that is, multiplying each feature with its corresponding initial weight, can further enhance the influence of important features while suppressing unimportant features.

[0092] The attention mechanism layer distributes weights so that the model can automatically screen out key features and pay more attention to them in subsequent processing, thereby improving the accuracy of fault detection. In fault detection, the self-attention mechanism can help the model understand the mutual influence between different features, so as to more accurately identify fault modes. Combining weight-based attention and self-attention mechanisms can also provide the model with a more comprehensive feature weighting capability. For example, the self-attention mechanism can be used to capture the dependencies within the features first, and then the weight-based attention can be used to further screen out key features. Therefore, the embodiments of the present invention can more comprehensively consider the importance and interrelationships of features, and further improve the accuracy and robustness of fault detection.

[0093] Next, the weighted fused electrical features and operating condition features are combined together to form a new feature vector. However, the linear combination may not be sufficient to capture the complex relationship between the features. Therefore, the present invention also performs a nonlinear transformation through an activation function, that is, it can introduce nonlinear factors to make the fused features more complex and diversified, thereby better representing the operating status and fault characteristics of the secondary protection panel cabinet.

[0094] The fusion features after nonlinear transformation not only contain the original information of electrical characteristics and operating condition characteristics, but also incorporate the interactions and relationships between them, which can support more accurate and reliable fault detection.

[0095] Therefore, the feature fusion layer distributes weights through the attention mechanism and generates multi-dimensional fusion features through linear combination and nonlinear transformation, which not only enhances the expressiveness of the features, but also improves the accuracy and robustness of fault detection.

[0096] Finally, the classification layer generates the fault type and fault warning degree of the secondary protection screen cabinet according to the fusion features, specifically:

[0097] The classification layer is used to calculate the similarity between the fusion feature and the weight of each fault type, convert the obtained similarity into probability, and output probability values ​​of several fault types;

[0098] The probability values ​​of various fault types are compared, the fault type with the highest probability value is taken as the fault type of the secondary protection panel cabinet, and the fault warning degree corresponding to the fault type with the highest probability value is output.

[0099] In schematic form, the classification layer can accurately measure the degree of match between the current equipment state and the known fault mode by calculating the similarity between the fusion features and the fault type weights, so as to reduce misjudgments and missed judgments and improve the accuracy of fault detection. Moreover, the classification layer can output the probability values ​​of multiple fault types at the same time, and is not limited to the judgment of a single fault, so it can more comprehensively evaluate the health status of the equipment and timely discover and deal with potential multiple fault risks.

[0100] Furthermore, by comparing the probability values ​​of various fault types, the fault type with the highest probability value, that is, the fault that is most likely to occur, can be prioritized, and the fault warning level corresponding to the fault type can be finally output, which can provide operation and maintenance personnel with important information about the severity of the fault, so that the operation and maintenance personnel can take corresponding response measures according to the urgency of the fault to avoid the expansion of the fault or more serious consequences.

[0101] For step S3, in a preferred embodiment, the fault warning information corresponding to the secondary protection screen cabinet is generated according to the fault type and the fault warning degree, specifically:

[0102] The classification results of the fault types are obtained from the classification layer to determine the fault types currently faced by the secondary protection panel cabinet, such as "air switch defect" and "loose wiring terminal", etc. The corresponding fault description can be obtained from the preset fault type library based on the classification results.

[0103] Furthermore, based on the quantitative value of the fault warning degree, such as a specific level (such as "low", "medium", "high", etc.), a value for indicating the urgency or warning degree of the fault can be obtained.

[0104] By integrating the fault type description and the warning degree description, complete fault warning information can be formed, and the fault warning information includes: fault description, current warning degree and possible impact scope or consequences.

[0105] Finally, the generated fault warning information is promptly delivered to the relevant operation and maintenance personnel, thereby providing them with timely and effective fault handling guidance.

[0106] like Figure 2 As shown, based on the above-mentioned various secondary protection screen cabinet fault detection method embodiments, the present invention provides a corresponding device embodiment;

[0107] An embodiment of the present invention provides a fault detection device for a secondary protection screen cabinet, comprising: a data acquisition module, a fault type determination module and an early warning module;

[0108] The data acquisition module is used to acquire the time series data and operating condition data corresponding to the secondary protection panel cabinet; wherein the time series data includes: voltage data and current data; the operating condition data includes: environmental index values ​​and operating parameters;

[0109] The fault type determination module is used to input the time series data and the operating condition data into a fault detection model, so that the fault detection model extracts the electrical features in the time series data and the operating condition features in the operating condition data, and fuses the electrical features with the operating condition features to generate multi-dimensional fusion features, and then generates the fault type and fault warning degree of the secondary protection panel cabinet according to the fusion features;

[0110] The warning module is used to generate fault warning information corresponding to the secondary protection screen cabinet according to the fault type and fault warning degree;

[0111] The generation of the fault detection model includes:

[0112] Each time series sample and each working condition sample of the secondary protection cabinet are used as corresponding training samples;

[0113] Taking each training sample, the actual fault type corresponding to each training sample and the actual fault warning degree corresponding to each training sample as input, and the predicted fault type of each training sample and the predicted fault warning degree of each training sample as output, the fault detection model to be trained is iteratively trained until the model converges to generate a fault detection model.

[0114] It should be noted that the device embodiments described above are merely schematic, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement it without paying creative labor.

[0115] Those skilled in the art can clearly understand that, for the sake of convenience and simplicity, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0116] Based on the above-mentioned various embodiments of the fault detection method for the secondary protection screen cabinet, the present invention provides corresponding embodiments of terminal equipment items.

[0117] An embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements a secondary protection screen cabinet fault detection method described in any method embodiment of the present invention.

[0118] The terminal device may be a computing terminal device such as a desktop computer, a notebook, a palm computer, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0119] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0120] The memory can be used to store the computer program, and the processor realizes various functions of the terminal device by running or executing the computer program stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device or other volatile solid-state storage device.

[0121] Based on the above-mentioned various embodiments of the fault detection method of the secondary protection screen cabinet, the present invention provides a corresponding storage medium item embodiment.

[0122] An embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute a fault detection method for a secondary protection screen cabinet described in any method embodiment of the present invention.

[0123] The storage medium is a computer-readable storage medium, and the computer program is stored in the computer-readable storage medium. When the computer program is executed by the processor, the steps of each method embodiment described above can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0124] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A fault detection method for a secondary protection panel cabinet, characterized in that: include: Obtaining the time series data and operating condition data corresponding to the secondary protection panel cabinet; wherein the time series data includes: voltage data and current data; the operating condition data includes: environmental index values ​​and operating parameters; Input the time series data and the operating condition data into a fault detection model, so that the fault detection model extracts the electrical features in the time series data and the operating condition features in the operating condition data, and fuses the electrical features with the operating condition features to generate multi-dimensional fusion features, and then generates the fault type and fault warning degree of the secondary protection panel cabinet according to the fusion features; Generate fault warning information corresponding to the secondary protection panel cabinet according to the fault type and fault warning degree; The generation of the fault detection model includes: Each time series sample and each working condition sample of the secondary protection cabinet are used as corresponding training samples; Taking each training sample, the actual fault type corresponding to each training sample and the actual fault warning degree corresponding to each training sample as input, and the predicted fault type of each training sample and the predicted fault warning degree of each training sample as output, the fault detection model to be trained is iteratively trained until the model converges to generate a fault detection model.

2. A fault detection method for a secondary protection screen cabinet as claimed in claim 1, characterized in that: The fault detection model includes: a feature extraction layer, a feature fusion layer and a classification layer; The feature extraction layer is used to extract electrical features based on voltage data and current data in the time series data, and to extract operating condition features in the operating condition data; The feature fusion layer is used to perform nonlinear transformation on the electrical features and the operating condition features to generate fusion features with multiple dimensions; The classification layer is used to generate the fault type and fault warning degree of the secondary protection screen cabinet according to the fusion characteristics.

3. A fault detection method for a secondary protection screen cabinet as claimed in claim 2, characterized in that: The working condition data also includes: communication parameters and attitude parameters; the communication parameters include: signal strength and data transmission rate; the attitude parameters include: device tilt angle and vibration amplitude; The feature extraction layer is used to extract electrical features based on voltage data and current data in the time series data, and to extract operating condition features in the operating condition data, including: The feature extraction layer performs temporal modeling on the time series data to capture long-term dependencies in the sequence; Through the long-term dependency in the sequence, a first hidden state sequence and a second hidden state sequence corresponding to the voltage data and the current data are output respectively; wherein the first hidden state sequence is used to represent the vector representation of the voltage in the time dimension; and the second hidden state sequence is used to represent the vector representation of the current in the time dimension; generating an electrical signature according to the first hidden state sequence and the second hidden state sequence; Based on the environmental index values, operating parameters, communication parameters, posture parameters, equipment tilt angles, and vibration amplitudes, operating condition characteristics for characterizing operating state changes of the secondary protection panel cabinet are extracted.

4. A fault detection method for a secondary protection screen cabinet as claimed in claim 3, characterized in that: The feature fusion layer is used to perform nonlinear transformation on the electrical features and the operating condition features to generate multi-dimensional fusion features, including: The feature fusion layer is used to assign different weights to the electrical features and the operating condition features, and then linearly combine the electrical features with the assigned weights and the operating condition features with the assigned weights; The activation function is used to perform nonlinear transformation on the features of the linear combination to generate multi-dimensional fusion features.

5. A fault detection method for a secondary protection screen cabinet as claimed in claim 4, characterized in that: The classification layer is used to generate the fault type and fault warning degree of the secondary protection screen cabinet according to the fusion feature, including: The classification layer is used to calculate the similarity between the fusion feature and the weight of each fault type, convert the obtained similarity into probability, and output probability values ​​of several fault types; The probability values ​​of various fault types are compared, the fault type with the highest probability value is taken as the fault type of the secondary protection panel cabinet, and the fault warning degree corresponding to the fault type with the highest probability value is output.

6. A fault detection method for a secondary protection screen cabinet as claimed in claim 5, characterized in that: The feature fusion layer is used to assign different weights to electrical features and operating condition features, including: The attention mechanism layer in the feature fusion layer calculates initial weights corresponding to the electrical features and the operating condition features, respectively, based on the input electrical features and the operating condition features; wherein the initial weights are used to characterize the contribution of the features to the fault type; The electrical characteristics and the operating condition characteristics are weighted and fused with the corresponding initial weights respectively to generate the weights corresponding to the electrical characteristics and the weights corresponding to the operating condition characteristics.

7. A fault detection method for a secondary protection screen cabinet as claimed in claim 1, characterized in that: In each iterative training, a training sample is input into the fault detection model, so that the fault detection model generates the predicted fault type and predicted fault warning degree corresponding to the training sample according to the electrical characteristics and working condition characteristics in the training sample; The predicted fault type and the predicted fault warning degree are compared with the actual fault type and the actual fault warning degree of the training sample, and then the network parameters of the fault detection model are adjusted according to the comparison results.

8. A fault detection device for a secondary protection panel cabinet, characterized in that: include: Data acquisition module, fault type determination module and early warning module; The data acquisition module is used to acquire the time series data and operating condition data corresponding to the secondary protection panel cabinet; wherein the time series data includes: voltage data and current data; the operating condition data includes: environmental index values ​​and operating parameters; The fault type determination module is used to input the time series data and the operating condition data into a fault detection model, so that the fault detection model extracts the electrical features in the time series data and the operating condition features in the operating condition data, and fuses the electrical features with the operating condition features to generate multi-dimensional fusion features, and then generates the fault type and fault warning degree of the secondary protection panel cabinet according to the fusion features; The warning module is used to generate fault warning information corresponding to the secondary protection screen cabinet according to the fault type and fault warning degree; The generation of the fault detection model includes: Each time series sample and each working condition sample of the secondary protection cabinet are used as corresponding training samples; Taking each training sample, the actual fault type corresponding to each training sample and the actual fault warning degree corresponding to each training sample as input, and the predicted fault type of each training sample and the predicted fault warning degree of each training sample as output, the fault detection model to be trained is iteratively trained until the model converges to generate a fault detection model.

9. A terminal device, characterized in that: It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements a fault detection method for a secondary protection screen cabinet as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute a fault detection method for a secondary protection screen cabinet as described in any one of claims 1 to 7.

Images