Fault arc detection method, apparatus, device, and readable storage medium
By using a support vector machine model trained with wavelet transform and particle swarm optimization algorithm to extract features and classify current signals, the problem of low accuracy in fault arc detection is solved, achieving higher detection accuracy and computational efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- INST OF ADVANCED TECH UNIV OF SCI & TECH OF CHINA
- Filing Date
- 2023-04-27
- Publication Date
- 2026-07-24
AI Technical Summary
Existing methods for detecting arc faults have low accuracy, especially in series circuits. Conventional line protection devices and arc fault detectors are easily affected by interference loads and environmental noise, leading to malfunctions and inaccurate detection.
Wavelet transform technology is used to process the current signal, extract high-frequency coefficients, and a support vector machine model is trained by particle swarm optimization algorithm to perform feature combination and classification, thereby establishing a fault arc detection model.
It improves the accuracy of fault arc detection, reduces computational costs, is highly adaptable, suitable for industrial embedded devices, and has practical engineering application value.
Smart Images
Figure CN116520082B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of arc detection technology, and in particular to a fault arc detection method, apparatus, equipment and readable storage medium. Background Technology
[0002] In low-voltage power supply lines and household appliances, fault arcs can be caused by aging or damage to wire insulation or loose electrical connections. Moreover, when a series arc fault occurs, the fault current is less than the normal current, making it impossible for conventional line protection devices (such as residual current circuit breakers and fuses) to make an accurate judgment on the arc fault.
[0003] Currently, fault arc detectors are commonly used to detect fault arcs. These detectors determine fault arcs based on the time-frequency domain characteristics of the current signal and by setting thresholds. However, the voltage and current of some interfering loads (such as dimmers, switching power supplies, motor loads, and equipment startup and shutdown) are very similar to those of series fault arcs. In addition, factors such as environmental noise, changes in load conditions, or network interference can easily cause fault arc detectors to malfunction, resulting in low accuracy of fault arc detection. Summary of the Invention
[0004] The main objective of this application is to provide a method, apparatus, device, and readable storage medium for detecting fault arcs, aiming to solve the technical problem of low accuracy in current fault arc detection.
[0005] To achieve the above objectives, this application provides a fault arc detection method, the fault arc detection method comprising:
[0006] Acquire the current signal of the circuit under test in a series circuit;
[0007] Perform wavelet transform on the current signal to obtain the corresponding high-frequency coefficients;
[0008] Based on the high-frequency coefficients, feature extraction and feature combination are performed on the current signal to obtain a feature matrix;
[0009] The feature matrix is input into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected. The fault arc detection model is a support vector machine model trained by the particle swarm optimization algorithm.
[0010] Optionally, the series circuit includes a current acquisition device and different types of loads. The current signal includes a normal current signal and a fault current signal. The different types of loads include at least one of a kettle and an incandescent lamp combination load, an air conditioner, a computer, and an incandescent lamp.
[0011] The steps for acquiring the current signal of the circuit under test in the series circuit include:
[0012] The current acquisition device collects the normal current signal and fault current signal of the combined load of the kettle and incandescent lamp, the air conditioner, the computer and the circuit to be tested corresponding to the incandescent lamp.
[0013] Optionally, the step of performing wavelet transform on the current signal to obtain the corresponding high-frequency coefficients includes:
[0014] Based on the line parameters of the line to be detected, a wavelet function is determined, wherein the line parameters include the line processing speed and the line storage capacity.
[0015] The number of decomposition layers is determined based on the current frequency distribution information of the current signal and the sampling frequency when the current signal is acquired;
[0016] Based on the wavelet function and the number of decomposition levels, wavelet transform is performed on each of the normal current signals and each of the fault current signals to obtain the corresponding high-frequency coefficients.
[0017] Optionally, the step of extracting and combining features from the current signal based on the high-frequency coefficients to obtain a feature matrix includes:
[0018] The high-frequency coefficients are input into each preset feature function to obtain each target feature, wherein the target features include waveform factor, impulse factor, margin factor, peak coefficient, range, peak-to-peak value, kurtosis coefficient, skewness and root mean square.
[0019] The target features are combined to obtain the feature matrix.
[0020] Optionally, the step of combining the target features to obtain the feature matrix includes:
[0021] Based on the number of decomposition layers, the maximum feature value of each target feature is selected;
[0022] Grouping the largest eigenvalues into multiple feature sets;
[0023] Calculate the standard deviation of the multiple feature sets, and combine the standard deviations to form the feature matrix.
[0024] Optionally, before the step of inputting the feature matrix into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected, the method further includes:
[0025] The feature matrix is divided into a training set and a test set, and the support vector machine model is trained based on the training set and the particle swarm optimization algorithm, and the optimal penalty parameters and optimal kernel function parameters of the support vector machine model are output.
[0026] Based on the optimal penalty parameters and the optimal kernel function parameters, the fault arc detection model is established.
[0027] Optionally, the step of inputting the feature matrix into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected includes:
[0028] The test set is input into the fault arc detection model, and the output data partitioning plane is divided into fault arc range and normal arc range, which includes the distribution points of all data in the test set.
[0029] If the distribution point is detected within the range of the fault arc, it is determined that a fault arc has occurred in the line under test.
[0030] Furthermore, to achieve the above objectives, this application also provides a fault arc detection device, the fault arc detection device comprising:
[0031] The current signal acquisition module is used to acquire the current signal of the circuit under test in a series circuit;
[0032] The wavelet transform module is used to perform wavelet transform on the current signal to obtain the corresponding high-frequency coefficients;
[0033] The feature matrix generation module is used to extract and combine features from the current signal based on the high-frequency coefficients to obtain a feature matrix;
[0034] The fault arc detection module is used to input the feature matrix into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected. The fault arc detection model is a support vector machine model trained by the particle swarm optimization algorithm.
[0035] Optionally, the current signal acquisition module is further used for:
[0036] The current acquisition device collects the normal current signal and fault current signal of the combined load of the kettle and incandescent lamp, the air conditioner, the computer and the circuit to be tested corresponding to the incandescent lamp.
[0037] Optionally, the wavelet transform module is further used for:
[0038] Based on the line parameters of the line to be detected, a wavelet function is determined, wherein the line parameters include the line processing speed and the line storage capacity.
[0039] The number of decomposition layers is determined based on the current frequency distribution information of the current signal and the sampling frequency when the current signal is acquired;
[0040] Based on the wavelet function and the number of decomposition levels, wavelet transform is performed on each of the normal current signals and each of the fault current signals to obtain the corresponding high-frequency coefficients.
[0041] Optionally, the feature matrix generation module is further configured to:
[0042] The high-frequency coefficients are input into each preset feature function to obtain each target feature, wherein the target features include waveform factor, impulse factor, margin factor, peak coefficient, range, peak-to-peak value, kurtosis coefficient, skewness and root mean square.
[0043] The target features are combined to obtain the feature matrix.
[0044] Optionally, the feature matrix generation module is further configured to:
[0045] Based on the number of decomposition layers, the maximum feature value of each target feature is selected;
[0046] Grouping the largest eigenvalues into multiple feature sets;
[0047] Calculate the standard deviation of the multiple feature sets, and combine the standard deviations to form the feature matrix.
[0048] Optionally, the fault arc detection device is further used for:
[0049] The feature matrix is divided into a training set and a test set, and the support vector machine model is trained based on the training set and the particle swarm optimization algorithm, and the optimal penalty parameters and optimal kernel function parameters of the support vector machine model are output.
[0050] Based on the optimal penalty parameters and the optimal kernel function parameters, the fault arc detection model is established.
[0051] Optionally, the fault arc detection module is further used for:
[0052] The test set is input into the fault arc detection model, and the output data partitioning plane is divided into fault arc range and normal arc range, which includes the distribution points of all data in the test set.
[0053] If the distribution point is detected within the range of the fault arc, it is determined that a fault arc has occurred in the line under test.
[0054] This application also provides a fault arc detection device, the fault arc detection comprising: a memory, a processor, and a fault arc detection program stored in the memory and executable on the processor, wherein when the fault arc detection program is executed by the processor, it implements the steps of the fault arc detection method described above.
[0055] This application also provides a readable storage medium storing a fault arc detection program, which, when executed by a processor, implements the steps of the fault arc detection method described above.
[0056] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the fault arc detection method described above.
[0057] This application provides a method, apparatus, device, and readable storage medium for detecting electric arc faults. Compared with current methods that use electric arc detectors for electric arc fault detection, this application first acquires the current signal of the circuit to be detected in a series circuit; performs wavelet transform on the current signal to obtain corresponding high-frequency coefficients. Since the high-frequency coefficients represent information about key changes in the current signal, using the high-frequency coefficients for subsequent feature extraction makes the extracted features more accurate and reliable. Moreover, wavelet transform can decorrelate the current signal, meaning that noise signals that cause interference are automatically separated and eliminated during the wavelet transform process, thus overcoming the technical problem of noise interference in current electric arc fault detection. Based on the high-frequency coefficients, feature extraction and feature combination are performed on the current signal to obtain a feature matrix; the feature matrix is input into an electric arc fault detection model to detect whether an electric arc fault has occurred in the circuit to be detected. The electric arc fault detection model refers to a Support Vector Machine (SVM) model trained using the Particle Swarm Optimization (PSO) algorithm. Support Vector Machine (SVM) models have extremely strong classification capabilities, accurately classifying faulty arcs. Furthermore, due to the strong generalization ability and fast convergence speed of the Particle Swarm Optimization (PSO) algorithm, the optimal parameters of the SVM model can be quickly and accurately found, improving the accuracy of the SVM model and thus enhancing the accuracy of faulty arc detection. Attached Figure Description
[0058] Figure 1 This is a flowchart illustrating the first embodiment of the fault arc detection method of this application;
[0059] Figure 2 This is a schematic diagram of the series circuit simulation platform in the fault arc detection method of this application;
[0060] Figure 3 This is a current waveform diagram involved in the fault arc detection method of this application;
[0061] Figure 4 This is a schematic diagram of the high-frequency coefficients involved in the fault arc detection method of this application;
[0062] Figure 5 This is a schematic diagram illustrating the accuracy of fault arc detection in the fault arc detection method of this application;
[0063] Figure 6 This is a schematic diagram comparing the actual labels and predicted labels involved in the fault arc detection method of this application;
[0064] Figure 7 This is a schematic diagram of the optimization process involved in the fault arc detection method of this application;
[0065] Figure 8 This is a schematic diagram of the device in the fault arc detection method of this application;
[0066] Figure 9 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the fault arc detection method of this application.
[0067] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0068] To make the above-mentioned objectives, features, and advantages of this application more apparent and understandable, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0069] In low-voltage power supply lines and household appliances, fault arcs can occur due to aging or damage to wire insulation or loose electrical connections. When a fault arc occurs, its core temperature can reach as high as 3000℃, and it radiates electrode material particles and hot gases outwards. Therefore, fault arcs are one of the main causes of electrical fires. In low-voltage AC power distribution systems, series fault arc detection is the most challenging. The current signal of a series fault arc is complex and closely related to the connected load. Therefore, this application focuses on fault arc detection for series circuits.
[0070] Conventional line protection devices, including residual current circuit breakers, fuses, or miniature circuit breakers, cannot accurately identify fault arcs when they occur because the current of the fault arc is less than the normal current. Therefore, those skilled in the art have gradually adopted fault arc detectors for fault arc detection. Fault arc detectors derive characteristic points from the fault arc's current signal through time-domain and frequency-domain analysis, and then use these characteristic points to set thresholds to identify the fault arc. However, the voltage and current of some interfering loads (such as dimmers, switching power supplies, motor loads, and equipment startup / shutdown) are very similar to those of a series fault arc. Combined with environmental noise, load condition variations, or network interference, these factors can easily cause false triggering of the fault arc detector, resulting in low accuracy in fault arc detection.
[0071] However, with the rapid development of artificial intelligence technology in recent years, more and more machine learning and deep learning network models are being applied to the field of arc fault detection. Although deep learning-based arc fault detection has achieved good results, it has high requirements for data samples, large computational overhead, and poor environmental adaptability.
[0072] To address the aforementioned issues, this application proposes processing current signals using wavelet transform and feature extraction, and then further classifying the current signals using a support vector machine model trained by particle swarm optimization, thereby distinguishing fault arcs. While ensuring detection accuracy, this approach effectively solves the problems of high data requirements, large computational overhead, and poor environmental adaptability associated with deep learning, thus reducing computational costs and facilitating deployment on industrial embedded devices, demonstrating practical value for real-world engineering applications.
[0073] This application provides a fault arc detection method. In the first embodiment of the fault arc detection method of this application, refer to... Figure 1 The fault arc detection method includes:
[0074] Step S10: Acquire the current signal of the circuit to be tested in the series circuit;
[0075] In this embodiment, it should be noted that the series circuit is applied to a pre-built AC series circuit experimental platform to simulate normal and fault arcs in actual power lines. (Refer to...) Figure 2As shown, the series circuit simulation platform includes a fault arc generating device, a current acquisition device, an AC power supply, an oscilloscope, a switch, and different types of loads. The fault arc generating device is used to simulate a fault arc. The current acquisition device includes a current transformer, a sampling resistor, and a data acquisition card, used to sense and acquire current signals. The AC power supply provides AC power to the current. The oscilloscope displays the current waveform of the acquired current signal in real time. The different types of loads include a combination load of a kettle and an incandescent lamp, an air conditioner, a computer, and at least one of incandescent lamps, used to simulate circuit faults under different operating conditions. The current signal includes normal current signals and fault current signals.
[0076] Step S10 includes:
[0077] Step S11: Collect the normal current signal and fault current signal of the combined load of the kettle and incandescent lamp, the air conditioner, the computer and the circuit to be tested corresponding to the incandescent lamp through the current acquisition device.
[0078] In this embodiment, it should be noted that, to more realistically simulate line faults under different operating conditions, loads of different power levels were selected, namely air conditioners, computers, and incandescent lamps. To enrich the collected current signals, this embodiment also considered collecting current signals from single loads and combined loads, namely the current signals from a kettle and an incandescent lamp combination. This makes the collected current signals closer to actual application scenarios and more convincing, thereby improving the reliability of fault arc detection. Those skilled in the art can select other types of loads according to actual needs, and can also select two or more combined loads, not limited to the types and combinations mentioned in this embodiment.
[0079] As an example, step S11 includes: when the current transformer senses the current signal of the line under test where different types of loads are located, the current signal is collected using a sampling resistor and a data acquisition card number. When a fault arc generator is installed on the line under test, the fault current signal is collected. When no fault arc generator is installed on the line under test, the normal current signal is collected. A total of eight sets of current data are collected. The sampling time interval can be set to 1 second. The eight sets of current data are divided into half a power frequency cycle, i.e., 0.01 seconds, to obtain the final current signal.
[0080] For example, when the load is an air conditioner, the current waveforms of the collected normal current signal and fault current signal are as follows: Figure 3 As shown in the figure, the horizontal axis represents the number of data points, i.e., the number of current signals collected, the vertical axis represents the amplitude, the dashed line represents the fault current signal, and the solid line represents the normal current signal.
[0081] Step S20: Perform wavelet transform on the current signal to obtain the corresponding high-frequency coefficients;
[0082] In this embodiment, it should be noted that wavelet transform is a new transform analysis method and an ideal tool for time-frequency analysis and processing of unstable signals, such as the current signal in this embodiment. Its main feature is that it can fully highlight certain characteristics through transformation. In addition, wavelet transform has the characteristics of high compression ratio and fast compression speed. After compression, it can keep the characteristics of the signal unchanged and resist interference during transmission. It can improve the accuracy of the characteristics of the current signal after transformation, thereby improving the accuracy of fault arc detection.
[0083] Step S20 includes:
[0084] Step S21: Determine the wavelet function based on the line parameters of the line to be detected, wherein the line parameters include the line processing speed and the line storage capacity;
[0085] Step S22: Determine the number of decomposition layers based on the current frequency distribution information of the current signal and the sampling frequency when acquiring the current signal;
[0086] Step S23: Based on the wavelet function and the number of decomposition layers, perform wavelet transform on each of the normal current signals and each of the fault current signals to obtain the corresponding high-frequency coefficients.
[0087] In this embodiment, it should be noted that: the line parameters are used to characterize the performance of the line under test, and may include the line's processing speed and storage capacity; the current frequency distribution information refers to the frequency distribution of the current signal, which is represented by the current waveform diagram mentioned above; the wavelet function is the operational form of wavelet transform, and may include classical wavelets such as Coiflet wavelets and Symlets wavelets, simple wavelets such as Haar wavelets, and orthogonal wavelets such as Meyer wavelets; the decomposition level refers to the number of decompositions performed when decomposing the current signal in wavelet transform.
[0088] As an example, steps S21 to S23 include: determining the wavelet function based on the operation rate and storage capacity of the circuit to be tested; considering that the classical wavelet function has a long filter and consumes a lot of resources, a simple wavelet function with a shorter filter length, namely the Haar wavelet, can be selected in this embodiment; based on the current frequency distribution information of the current signal, the complexity of the current signal can be analyzed. The more complex the signal, the more decompositions are required. Combined with the sampling frequency when acquiring the current signal, the final decomposition level of the current signal is determined. In this embodiment, four-level decomposition is selected; based on the wavelet function and the number of decomposition levels, wavelet transform is performed on each normal current signal and each fault current signal to obtain the corresponding high-frequency coefficients.
[0089] For example, such as Figure 4 The diagram shows the high-frequency coefficients obtained after performing four-level wavelet transforms on the normal current signal and fault current signal of the collected air conditioner. The horizontal axis represents the number of current signals in each decomposition, and the vertical axis represents the high-frequency coefficients.
[0090] Because the accuracy of wavelet transform is closely related to the choice of wavelet function, and different filters correspond to different wavelet functions, the filter is affected to some extent by the line's computing speed and storage capacity. Therefore, in this embodiment, a suitable wavelet function is determined based on the line's computing power and storage capacity. Secondly, the accuracy of wavelet transform is also related to the choice of the number of decomposition layers. Fewer decomposition layers will reduce the resolution, while more decomposition layers will slow down the analysis speed. Therefore, in this embodiment, the complexity of the current signal is determined based on the current frequency distribution information, and a suitable number of decomposition layers is determined in combination with the sampling frequency, thereby completing the wavelet transform of each current signal, improving the accuracy of the wavelet transform, thereby improving the accuracy of subsequent feature extraction, and ultimately improving the accuracy of the entire fault arc detection.
[0091] Step S30: Based on the high-frequency coefficients, perform feature extraction and feature combination on the current signal to obtain a feature matrix;
[0092] In this embodiment, it should be noted that the feature extraction refers to inputting the high-frequency coefficients into each preset feature function to obtain each target feature; the feature combination refers to filtering and grouping each target feature to form a feature matrix; the feature matrix is used as the input of the subsequent fault arc detection model.
[0093] Step S30 includes:
[0094] Step S31: Input the high-frequency coefficients into each preset feature function to obtain each target feature, wherein the target features include waveform factor, impulse factor, margin factor, peak coefficient, range, peak-to-peak value, kurtosis coefficient, skewness and root mean square.
[0095] Step S32: Combine the target features to obtain the feature matrix.
[0096] In this embodiment, it should be noted that, considering that it is difficult to fully capture fault arcs with a single feature, nine features were selected: waveform factor, impulse factor, margin factor, peak coefficient, range, peak-to-peak value, kurtosis coefficient, skewness, and root mean square. The preset feature function refers to the preset operation function of the target feature, which includes waveform factor function, impulse factor function, margin factor function, peak coefficient function, range function, peak-to-peak value function, kurtosis coefficient function, skewness function, and root mean square function.
[0097] As an example, steps S31 to S32 include: inputting the high-frequency coefficients of the normal current signal and the fault current signal at each decomposition level into each preset feature function to obtain multiple target features, wherein the target features include waveform factor, pulse factor, margin factor, peak coefficient, range, peak-to-peak value, kurtosis coefficient, skewness and root mean square.
[0098] The waveform factor function is:
[0099] The impulse factor function is:
[0100] The margin factor function is:
[0101] The peak coefficient function is:
[0102]
[0103] The range function is: max(|x i |)-min(|x i |)
[0104] The peak-to-peak value function is: |max(x) i )|+|min(x i )|
[0105] The kurtosis coefficient function is:
[0106] The skewness function is:
[0107] The root mean square function is:
[0108] Where x represents the high-frequency coefficient and n represents the number of current signals;
[0109] After calculating each waveform factor, impulse factor, margin factor, peak value coefficient, range, peak-to-peak value, kurtosis coefficient, skewness, and root mean square, the features are combined to obtain the feature matrix.
[0110] Step S32 includes:
[0111] Step S321: Based on the number of decomposition layers, select the maximum feature value of each target feature;
[0112] Step S322: Group the maximum eigenvalues to obtain multiple feature sets;
[0113] Step S323: Calculate the standard deviation of the multiple feature sets, and combine the standard deviations to form the feature matrix.
[0114] In this embodiment, it should be noted that since the influence of electric arcs varies in different frequency bands of each decomposition layer, this embodiment uses the maximum value of each target feature as a new feature for feature combination to amplify the existence of electric arcs. This not only allows for better capture of the faulty electric arcs that have an impact, but also effectively reduces the amount of computation, thereby improving the efficiency of feature combination and thus improving the accuracy and efficiency of subsequent faulty electric arc detection.
[0115] As an example, steps S321 to S323 include: normal current signals and fault current signals have corresponding high-frequency coefficients in each decomposition layer, that is, there are corresponding waveform factors, impulse factors, margin factors, peak coefficients, ranges, peak-to-peak values, kurtosis coefficients, skewness, and root mean square. The largest eigenvalue of these features in the four-layer decomposition is selected, and then each largest eigenvalue is grouped according to a preset grouping rule. The preset grouping rule can be how many eigenvalues a group contains, or other grouping rules can be set according to the actual situation to obtain multiple feature sets. Then, the standard deviation of each feature set is calculated to form a feature matrix.
[0116] Step S40: Input the feature matrix into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected, wherein the fault arc detection model refers to a support vector machine model trained by the particle swarm optimization algorithm.
[0117] In this embodiment, it should be noted that the particle swarm optimization algorithm is a search algorithm based on group cooperation, developed by simulating the foraging behavior of bird flocks. It belongs to heuristic algorithms, also known as intelligent optimization algorithms. Its basic idea is to find the optimal solution through cooperation and information sharing among individuals in the group. The support vector machine model is a binary classification model. Its basic model is a linear classifier with the largest margin defined on the feature space. However, data classification using the support vector machine model requires determining two parameters: the penalty parameter and the kernel function parameter. Therefore, the particle swarm optimization algorithm is used to train the support vector machine model to calculate the optimal penalty parameter and the optimal kernel function parameter, thereby improving the accuracy of the support vector machine model and thus improving the accuracy of fault arc detection.
[0118] Step S40 includes:
[0119] Step S41: Input the test set into the fault arc detection model and output a data partitioning plane, wherein the data partitioning plane is divided into a fault arc range and a normal arc range, and includes the distribution points of all data in the test set;
[0120] Step S42: If the distribution point is detected within the range of the fault arc, it is determined that a fault arc has occurred in the line to be tested.
[0121] As an example, steps S41 to S42 include: inputting the test set into the fault arc detection model to classify fault current signals and normal current signals, and outputting a data partitioning plane, wherein the data partitioning plane is divided into a fault arc range and a normal arc range, containing the distribution points of all data in the test set, that is, the distribution points of all normal current signals and all fault current signals; if a distribution point is detected within the fault arc range, it indicates that a fault arc exists on the line to be detected. At this time, it can be further detected whether the distribution points within the fault arc range are consistent with the previously collected fault current signals to verify the accuracy of the fault arc detection model.
[0122] For example, taking an air conditioner as the load, the accuracy of the fault arc detection model in detecting the fault arc in the circuit where the air conditioner is located is as follows: Figure 5 As shown in the figure, the horizontal axis represents the number of model training iterations, and the vertical axis represents the accuracy. It can be seen that the accuracy reaches 99.697%. Figure 6 This is a diagram showing the comparison between the collected real labels and the predicted labels calculated by the model.
[0123] Prior to step S40, the following steps are also included:
[0124] Step A10: Divide the feature matrix into a training set and a test set, and train the support vector machine model based on the training set and the particle swarm optimization algorithm, and output the optimal penalty parameters and optimal kernel function parameters of the support vector machine model;
[0125] Step A20: Based on the optimal penalty parameters and the optimal kernel function parameters, establish the fault arc detection model.
[0126] In this embodiment, it should be noted that since the collected normal current signal and fault current signal are linearly inseparable, a Gaussian kernel is introduced to map the training set to a high-dimensional space, making the data linearly separable in the high-dimensional space. The kernel function corresponding to the Gaussian kernel is the radial basis function, and its expression is: K(x1,x2)=exp(-g||x1-x2|| 2 ), g>0, where K represents the kernel function, x represents the data point, and g represents the kernel function parameter; the penalty parameter is the decision parameter of the support vector machine model. If the penalty parameter is too large, it will lead to overfitting, and if it is too small, it will lead to underfitting. Therefore, selecting the optimal penalty parameter can effectively improve the accuracy of the fault arc detection model.
[0127] As an example, steps A10 to A20 include: dividing the feature matrix into a training set and a training set according to a preset ratio (in this embodiment, the ratio is 7:3); inputting the training set into the support vector machine model; continuously training the model using the particle swarm optimization algorithm to find the optimal solution; outputting the current optimization result; and constructing a two-dimensional parameter optimization space for the particles using penalty parameters and kernel function parameters. Selecting an optimal solution from this space, namely the optimal penalty parameters and the optimal kernel function parameters, is then performed. The optimization process can be referred to... Figure 7 As shown: First, determine the fitness function for particle swarm optimization, using the recognition rate of the test set as the fitness function. Employ cross-validation to initialize the population and set particle swarm parameters. The fitness value is taken as the average classification accuracy. Then, calculate the fitness value of each particle as the individual extreme value of the current particle, and the largest individual extreme value as the global extreme value. Next, update the particle's position and velocity, and determine whether the current state meets the termination condition. If it does, the optimization ends; otherwise, continue updating the particle's position and velocity. Finally, update the individual extreme value and the global extreme value until the optimization ends. Output the current optimization result, the penalty coefficient c, and the kernel function parameter g to form the two-dimensional parameter optimization space of the particle, from which the optimal solution is selected. Input the optimal penalty parameter and the optimal kernel function parameter into the support vector machine model to obtain the fault arc detection model.
[0128] This embodiment provides a fault arc detection method. Compared with the current method of using fault arc detectors for fault arc detection, this embodiment first acquires the current signal of the circuit to be detected in the series circuit; performs wavelet transform on the current signal to obtain the corresponding high-frequency coefficients. Since the high-frequency coefficients represent the information when the current signal changes significantly, using the high-frequency coefficients for subsequent feature extraction makes the extracted features more accurate and reliable. Moreover, wavelet transform can decorrelate the current signal, that is, during the wavelet transform process, the noise signal that generates interference is automatically separated and eliminated, thereby overcoming the technical problem of noise interference in current fault arc detection. Based on the high-frequency coefficients, feature extraction and feature combination are performed on the current signal to obtain a feature matrix; the feature matrix is input into a fault arc detection model to detect whether a fault arc has occurred in the circuit to be detected. The fault arc detection model refers to a support vector machine (SVM) model trained by the particle swarm optimization (PSO) algorithm. Support Vector Machine (SVM) models have extremely strong classification capabilities, accurately classifying faulty arcs. Furthermore, due to the strong generalization ability and fast convergence speed of the Particle Swarm Optimization (PSO) algorithm, the optimal parameters of the SVM model can be quickly and accurately found, improving the accuracy of the SVM model and thus enhancing the accuracy of faulty arc detection.
[0129] Furthermore, embodiments of this application also provide a fault arc detection device, such as... Figure 8 As shown, the fault arc detection device includes:
[0130] The current signal acquisition module 10 is used to acquire the current signal of the circuit to be tested in the series circuit;
[0131] Wavelet transform module 20 is used to perform wavelet transform on the current signal to obtain the corresponding high-frequency coefficients;
[0132] The feature matrix generation module 30 is used to extract and combine features of the current signal based on the high-frequency coefficients to obtain a feature matrix;
[0133] The fault arc detection module 40 is used to input the feature matrix into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected. The fault arc detection model refers to a support vector machine model trained by the particle swarm optimization algorithm.
[0134] Optionally, the current signal acquisition module 10 is further used for:
[0135] The current acquisition device collects the normal current signal and fault current signal of the combined load of the kettle and incandescent lamp, the air conditioner, the computer and the circuit to be tested corresponding to the incandescent lamp.
[0136] Optionally, the wavelet transform module 20 is further configured to:
[0137] Based on the line parameters of the line to be detected, a wavelet function is determined, wherein the line parameters include the line processing speed and the line storage capacity.
[0138] The number of decomposition layers is determined based on the current frequency distribution information of the current signal and the sampling frequency when the current signal is acquired;
[0139] Based on the wavelet function and the number of decomposition levels, wavelet transform is performed on each of the normal current signals and each of the fault current signals to obtain the corresponding high-frequency coefficients.
[0140] Optionally, the feature matrix generation module 30 is further configured to:
[0141] The high-frequency coefficients are input into each preset feature function to obtain each target feature, wherein the target features include waveform factor, impulse factor, margin factor, peak coefficient, range, peak-to-peak value, kurtosis coefficient, skewness and root mean square.
[0142] The target features are combined to obtain the feature matrix.
[0143] Optionally, the feature matrix generation module 30 is further configured to:
[0144] Based on the number of decomposition layers, the maximum feature value of each target feature is selected;
[0145] Grouping the largest eigenvalues into multiple feature sets;
[0146] Calculate the standard deviation of the multiple feature sets, and combine the standard deviations to form the feature matrix.
[0147] Optionally, the fault arc detection device is further used for:
[0148] The feature matrix is divided into a training set and a test set, and the support vector machine model is trained based on the training set and the particle swarm optimization algorithm, and the optimal penalty parameters and optimal kernel function parameters of the support vector machine model are output.
[0149] Based on the optimal penalty parameters and the optimal kernel function parameters, the fault arc detection model is established.
[0150] Optionally, the fault arc detection module 40 is further used for:
[0151] The test set is input into the fault arc detection model, and the output data partitioning plane is divided into fault arc range and normal arc range, which includes the distribution points of all data in the test set.
[0152] If the distribution point is detected within the range of the fault arc, it is determined that a fault arc has occurred in the line under test.
[0153] The fault arc detection device provided in this application employs the fault arc detection method in the above embodiments, solving the technical problem of low accuracy in current fault arc detection. Compared with the prior art, the beneficial effects of the fault arc detection device provided in this application are the same as those of the fault arc detection method provided in the above embodiments, and other technical features in this fault arc detection device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.
[0154] This application provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the fault arc detection method in Embodiment 1 above.
[0155] The following is for reference. Figure 9 The diagram illustrates a structural schematic of an electronic device suitable for implementing embodiments of the present disclosure. The electronic devices in the embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Android Devices), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 9 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.
[0156] like Figure 9As shown, an electronic device may include a processing unit (such as a central processing unit, graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) or a program loaded from a storage device into random access memory (RAM). The RAM also stores various programs and data required for the operation of the electronic device. The processing unit, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.
[0157] Typically, the following systems can be connected to the I / O interface: input devices such as touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices such as liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices such as magnetic tapes, hard drives, etc.; and communication devices. Communication devices allow electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although electronic devices with various systems are shown in the figures, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems may be implemented alternatively.
[0158] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processing device, it performs the functions defined above in the methods of embodiments of this disclosure.
[0159] The electronic device provided in this application employs the fault arc detection method in the above embodiments, solving the technical problem of low accuracy in current fault arc detection. Compared with the prior art, the beneficial effects of the electronic device provided in this application are the same as those of the fault arc detection method provided in the above embodiments, and other technical features of the electronic device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.
[0160] It should be understood that various parts of this disclosure can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.
[0161] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0162] This embodiment provides a computer-readable storage medium having computer-readable program instructions stored thereon, which are used to execute the fault arc detection method in Embodiment 1 above.
[0163] The computer-readable storage medium provided in this application embodiment may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, 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 thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.
[0164] The aforementioned computer-readable storage medium may be included in an electronic device or may exist independently without being assembled into an electronic device.
[0165] The aforementioned computer-readable storage medium carries one or more programs. When the aforementioned one or more programs are executed by an electronic device, the electronic device causes the electronic device to: acquire the current signal of the circuit to be detected in the series circuit; perform wavelet transform on the current signal to obtain the corresponding high-frequency coefficients; perform feature extraction and feature combination on the current signal based on the high-frequency coefficients to obtain a feature matrix; and input the feature matrix into a fault arc detection model to detect whether a fault arc has occurred in the circuit to be detected, wherein the fault arc detection model refers to a support vector machine model trained by the particle swarm optimization algorithm.
[0166] Computer program code for performing the operations of this disclosure can be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, and C++, and conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0167] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.
[0168] The modules described in the embodiments of this disclosure can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.
[0169] The computer-readable storage medium provided in this application stores computer-readable program instructions for executing the above-described fault arc detection method, thus solving the technical problem of low accuracy in current fault arc detection methods. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the fault arc detection method provided in the above-described embodiments, and will not be repeated here.
[0170] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the fault arc detection method described above.
[0171] The computer program product provided in this application solves the technical problem of low accuracy in current fault arc detection. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the fault arc detection method provided in the above embodiments, and will not be repeated here.
[0172] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent scope of this application.
Claims
1. A method for detecting fault arcs, characterized in that, The fault arc detection method includes: Acquire the current signal of the circuit under test in a series circuit; Perform wavelet transform on the current signal to obtain the corresponding high-frequency coefficients; Based on the high-frequency coefficients, feature extraction and feature combination are performed on the current signal to obtain a feature matrix; The feature matrix is input into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected. The fault arc detection model is a support vector machine model trained by the particle swarm optimization algorithm. The step of extracting and combining features from the current signal based on the high-frequency coefficients to obtain a feature matrix includes: The high-frequency coefficients are input into each preset feature function to obtain each target feature, wherein the target features include waveform factor, impulse factor, margin factor, peak coefficient, range, peak-to-peak value, kurtosis coefficient, skewness and root mean square. Based on the number of decomposition layers, the maximum eigenvalue of each target feature is selected, and each maximum eigenvalue is grouped to obtain multiple feature sets. The standard deviation of the multiple feature sets is calculated, and the standard deviations are combined to form the feature matrix.
2. The fault arc detection method as described in claim 1, characterized in that, The series circuit includes a current acquisition device and different types of loads. The current signal includes a normal current signal and a fault current signal. The different types of loads include at least one of the following: a combination load of a kettle and an incandescent lamp, an air conditioner, a computer, and an incandescent lamp. The steps for acquiring the current signal of the circuit under test in the series circuit include: The current acquisition device collects the normal current signal and fault current signal of the combined load of the kettle and incandescent lamp, the air conditioner, the computer and the circuit to be tested corresponding to the incandescent lamp.
3. The fault arc detection method as described in claim 2, characterized in that, The step of performing wavelet transform on the current signal to obtain the corresponding high-frequency coefficients includes: Based on the line parameters of the line to be detected, a wavelet function is determined, wherein the line parameters include the line processing speed and the line storage capacity. The number of decomposition layers is determined based on the current frequency distribution information of the current signal and the sampling frequency when the current signal is acquired; Based on the wavelet function and the number of decomposition levels, wavelet transform is performed on each of the normal current signals and each of the fault current signals to obtain the corresponding high-frequency coefficients.
4. The fault arc detection method as described in claim 1, characterized in that, Before the step of inputting the feature matrix into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected, the method further includes: The feature matrix is divided into a training set and a test set, and the support vector machine model is trained based on the training set and the particle swarm optimization algorithm, and the optimal penalty parameters and optimal kernel function parameters of the support vector machine model are output. Based on the optimal penalty parameters and the optimal kernel function parameters, the fault arc detection model is established.
5. The fault arc detection method as described in claim 4, characterized in that, The step of inputting the feature matrix into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected includes: The test set is input into the fault arc detection model, and the output data partitioning plane is divided into fault arc range and normal arc range, which includes the distribution points of all data in the test set. If the distribution point is detected within the range of the fault arc, it is determined that a fault arc has occurred in the line under test.
6. A fault arc detection device, characterized in that, The fault arc detection device includes: The current signal acquisition module is used to acquire the current signal of the circuit under test in a series circuit; The wavelet transform module is used to perform wavelet transform on the current signal to obtain the corresponding high-frequency coefficients; The feature matrix generation module is used to extract and combine features from the current signal based on the high-frequency coefficients to obtain a feature matrix; The fault arc detection module is used to input the feature matrix into the fault arc detection model to detect whether a fault arc has occurred in the line to be detected, wherein the fault arc detection model refers to a support vector machine model trained by the particle swarm optimization algorithm. The feature matrix generation module is further configured to input the high-frequency coefficients into each preset feature function to obtain each target feature, wherein the target features include waveform factor, impulse factor, margin factor, peak coefficient, range, peak-to-peak value, kurtosis coefficient, skewness, and root mean square; based on the number of decomposition layers, the maximum eigenvalue of each target feature is selected, and each maximum eigenvalue is grouped to obtain multiple feature sets, the standard deviation of the multiple feature sets is calculated, and the standard deviations are combined to form the feature matrix.
7. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the steps of the fault arc detection method according to any one of claims 1 to 5.
8. A readable storage medium, characterized in that, The readable storage medium stores a program for implementing a fault arc detection method, which is executed by a processor to implement the steps of the fault arc detection method as described in any one of claims 1 to 5.