Power quality interference signal type detection method, detection device and storage medium

Through the combination of wavelet packet transformation and ELM model, denoising and feature extraction of the power quality interference signal is solved, which solves the problem of inaccurate type detection caused by noise interference, and achieves more efficient power quality interference signal type recognition.

CN119089242APending Publication Date: 2024-12-06GUANGDONG POWER GRID CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411131038.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

In the prior art, noise in the power quality interference signal causes inaccurate type detection and lacks effective solutions.

Method used

The wavelet packet transformation method is used to denoise the original power quality interference signal. By obtaining the sampling frequency and fundamental frequency, dividing the frequency band number, and performing denoising processing, the target power quality interference signal is obtained, and the feature vector is extracted using the ELM model and the LMD algorithm. , identify the signal type.

Benefits of technology

It improves the accuracy of signal type detection of power quality interference, reduces noise interference, and achieves more accurate signal type recognition.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089242A_ABST
    Figure CN119089242A_ABST
Patent Text Reader

Abstract

The invention provides a power quality interference signal type detection method and device and a storage medium. The method comprises the following steps: carrying out de-noising processing on an original power quality interference signal by adopting a wavelet packet transformation method to obtain a target power quality interference signal; and determining the type of the original power quality interference signal according to the target power quality interference signal. According to the method, the problem of inaccurate detection of the type of the power quality interference signal caused by noise in the power quality interference signal in the prior art is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of power quality interference signal type detection, and in particular to a power quality interference signal type detection method, a power quality interference signal type detection device, a computer-readable storage medium and a computer program product. Background Art

[0002] In the power quality interference signal type detection method in the prior art, the accuracy of type detection is greatly affected by the noise in the power quality interference signal, and there is a problem that the noise in the power quality interference signal leads to inaccurate power quality interference signal type detection.

[0003] Currently, there is no solution to the above problem. Summary of the invention

[0004] The main purpose of the present application is to provide a method for detecting the type of a power quality interference signal, a device for detecting the type of a power quality interference signal, a computer-readable storage medium and a computer program product, so as to at least solve the problem in the prior art that the noise in the power quality interference signal causes inaccurate detection of the type of the power quality interference signal.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a method for detecting the type of power quality interference signal is provided, the method comprising: using a wavelet packet transform method to denoise the original power quality interference signal to obtain a target power quality interference signal; and determining the type of the original power quality interference signal according to the target power quality interference signal.

[0006] Optionally, the original power quality interference signal is denoised by a wavelet packet transform method to obtain a target power quality interference signal, including: obtaining a sampling frequency of the original power quality interference signal and a base frequency of the power quality interference signal; and Determine the number of divided frequency bands, where K is the number of divided frequency bands, f s is the sampling frequency, f is the base frequency; based on the number of divided frequency bands, the wavelet packet transform method is used to denoise the original power quality interference signal to obtain the target power quality interference signal.

[0007] Optionally, determining the type of the original power quality interference signal according to the target power quality interference signal includes: performing feature extraction on the target power quality interference signal to obtain multiple target feature vectors, wherein the target feature vector is a feature vector of the target power quality interference signal; inputting the multiple target feature vectors into a recognition model to obtain multiple target probabilities, wherein the recognition model is obtained by training an ELM model using multiple groups of training data, each group of training data at least includes: multiple feature vectors of the power quality interference signal, multiple predicted probabilities of the power quality interference signal acquired during a historical period, one predicted probability corresponds to a predetermined type, the predicted probability is the probability that the type is the predetermined type corresponding to the predicted probability, and the target probability is the predicted probability of the target power quality interference signal; determining that the predetermined type corresponding to the maximum target probability is the type of the original power quality interference signal.

[0008] Optionally, feature extraction is performed on the target power quality interference signal to obtain multiple target feature vectors, including: using an LMD algorithm to decompose the target power quality interference signal to obtain multiple target feature vectors, where the target feature vectors are PF components.

[0009] Optionally, multiple target feature vectors are input into a recognition model to obtain multiple target probabilities, including: inputting multiple target feature vectors into an input layer, the input layer is composed of multiple input layer neurons, and one input layer neuron is used to receive one target feature vector; inputting the target feature vector received by the input layer into a hidden layer to obtain an output matrix, the hidden layer is composed of multiple hidden layer neurons; inputting the output matrix into an output layer to obtain a probability vector, the output layer is composed of multiple output layer neurons, and the probability vector contains all the target probabilities.

[0010] Optionally, the number of input layer neurons is the same as the number of target probabilities, the number of hidden layer neurons is twice the number of target probabilities, the number of output layer neurons is the same as the number of target probabilities, and the bias of the hidden layer neurons is twice the number of target probabilities.

[0011] Optionally, the types include: normal voltage, voltage fluctuation, voltage instability, voltage interruption, instantaneous impulse, and transient oscillation.

[0012] According to another aspect of the present application, a device for detecting the type of a power quality interference signal is provided, the device comprising: a denoising unit, used to perform denoising processing on an original power quality interference signal using a wavelet packet transform method to obtain a target power quality interference signal; and a determination unit, used to determine the type of the original power quality interference signal based on the target power quality interference signal.

[0013] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute any one of the power quality interference signal type detection methods.

[0014] According to another aspect of the present application, a computer program product is provided, comprising a computer program / instruction, which implements any one of the power quality interference signal type detection methods when executed by a processor.

[0015] By applying the technical solution of the present application, first, the original power quality interference signal is denoised using a wavelet packet transform method to obtain a target power quality interference signal, and then the type of the original power quality interference signal is determined based on the target power quality interference signal, thereby solving the problem of inaccurate detection of the type of power quality interference signal caused by noise in the power quality interference signal in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A hardware structure block diagram of a mobile terminal for executing a method for detecting a power quality interference signal type provided in an embodiment of the present application is shown;

[0017] Figure 2 A schematic diagram of a flow chart of a method for detecting a type of power quality interference signal provided according to an embodiment of the present application is shown;

[0018] Figure 3 A schematic diagram of an ELM model provided according to an embodiment of the present application is shown;

[0019] Figure 4 A schematic diagram showing a signal-to-noise ratio of a different denoising method provided according to an embodiment of the present application;

[0020] Figure 5 A schematic diagram showing a mean square error of a different denoising method provided according to an embodiment of the present application;

[0021] Figure 6 A structural block diagram of a power quality interference signal type detection device provided according to an embodiment of the present application is shown.

[0022] The above drawings include the following reference numerals:

[0023] 102, processor; 104, memory; 106, transmission device; 108, input and output devices. DETAILED DESCRIPTION

[0024] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0025] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0026] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0027] For the convenience of description, some nouns or terms involved in the embodiments of the present application are explained below:

[0028] Wavelet Packet Transform (WPT) is a multi-scale analysis method that can decompose signals into sub-signals of different frequency bands. Wavelet Packet Transform (WPT) is a signal processing technology based on wavelet transform that can decompose and reconstruct signals. In denoising applications, WPT can decompose signals into sub-bands of different frequencies, process each sub-band, and then merge the processed sub-bands into the original signal through reconstruction.

[0029] Number of frequency band divisions: refers to the number of frequency bands into which the signal is decomposed when performing wavelet packet decomposition. More frequency band divisions mean a more detailed decomposition of the signal, which can better capture the detailed information of the signal, but it will also increase the computational complexity.

[0030] LMD algorithm (Local Moran's I, local mean decomposition method): is a statistical method used for spatial data analysis, often used to study spatial autocorrelation in geographic data. The algorithm is based on Moran's I statistics, but takes local spatial relationships into account during the calculation process in order to better understand the local characteristics of spatial data.

[0031] Singular Value Decomposition (SVD) denoising: It is a matrix decomposition-based method that removes noise by retaining the most important singular values ​​in the data matrix. This method is widely used in image processing and signal processing.

[0032] Wavelet adaptive threshold denoising: It is a denoising method based on wavelet transform. It adaptively selects the threshold according to the characteristics of the signal, thus achieving better denoising effect.

[0033] Wavelet soft threshold denoising: It is a denoising method based on wavelet transform. It performs soft threshold processing on the wavelet coefficients of the signal and sets the coefficients less than the threshold to 0, thereby achieving the denoising effect. This method is also widely used in signal processing and image processing.

[0034] ELM model (Extreme Learning Machine): is a machine learning model based on a single-layer neural network. Unlike traditional neural network models, the weights between the input layer and the hidden layer of the ELM model are randomly initialized, while the weights between the output layer and the hidden layer are directly calculated by analytical methods. This simple structure gives the ELM model advantages in training speed and generalization ability, and is particularly suitable for processing large-scale data sets and high-dimensional data. The ELM model is usually used in machine learning tasks such as classification, regression, and clustering.

[0035] SVM model (Support Vector Machine): is a supervised learning algorithm used for classification and regression analysis. It separates samples of different categories by finding an optimal hyperplane to maximize the interval between samples. The SVM model performs well when processing high-dimensional data and nonlinear data.

[0036] BPNN model (Back Propagation Neural Network): is a supervised learning algorithm based on neural network. It continuously adjusts the weight parameters in the network through the back propagation algorithm to minimize the error between the predicted value and the actual value. The BPNN model can be used for classification, regression and other complex data analysis tasks.

[0037] As introduced in the background technology, the noise in the power quality interference signal in the prior art leads to inaccurate detection of the power quality interference signal type. In order to solve the problem of inaccurate detection of the power quality interference signal type due to the noise in the power quality interference signal in the prior art, the embodiments of the present application provide a power quality interference signal type detection method, a power quality interference signal type detection device, a computer-readable storage medium and a computer program product.

[0038] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0039] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 FIG. 1 is a hardware structure block diagram of a mobile terminal of a method for detecting a type of power quality interference signal according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0040] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the display method of device information in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. The transmission device 106 is used to receive or send data via a network. The above-mentioned specific examples of the network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0041] In this embodiment, a method for detecting the type of power quality interference signal running on a mobile terminal, a computer terminal or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0042] Figure 2 1 is a flow chart of a method for detecting a power quality interference signal type according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0043] Step S201, using a wavelet packet transform method to perform denoising on the original power quality interference signal to obtain a target power quality interference signal;

[0044] In an optional solution, the above step S201 can be implemented as follows:

[0045] Obtaining the sampling frequency of the original power quality interference signal and the base frequency of the power quality interference signal;

[0046] According to the above sampling frequency, the above base frequency and Determine the number of divided frequency bands, where K is the number of divided frequency bands, fs is the above sampling frequency, f is the above base frequency;

[0047] Based on the above-mentioned number of divided frequency bands, the above-mentioned wavelet packet transform method is used to perform denoising processing on the above-mentioned original power quality interference signal to obtain the above-mentioned target power quality interference signal.

[0048] Specifically, in the denoising process of the power quality interference signal, the number of frequency band divisions must be determined. The selection of the number of frequency band divisions has an important influence on the denoising result. If the number of frequency band divisions is too small, the denoising effect may not be ideal and the signal-to-noise ratio can only be slightly improved. Selecting too many frequency band divisions may cause information loss. Therefore, it is very important to accurately determine the number of frequency band divisions. The principles for determining the number of frequency band divisions are as follows: Ensure that the fundamental frequency component of the power quality interference signal is located at the center of the lowest sub-band to limit its impact on other sub-bands. The actual number of frequency band divisions of the power quality interference signal is one less than the number of sub-bands. Using the formula The calculated value can be rounded down to obtain a reasonable number of frequency band divisions.

[0049] Step S202: determining the type of the original power quality interference signal according to the target power quality interference signal.

[0050] In an optional solution, the above step S202 can be implemented as follows:

[0051] Step S2021, extracting features of the target power quality interference signal to obtain a plurality of target feature vectors, where the target feature vectors are feature vectors of the target power quality interference signal;

[0052] In an optional solution, the above step S2021 can be implemented as follows:

[0053] The target power quality interference signal is decomposed and processed by using the LMD algorithm to obtain a plurality of target feature vectors, where the target feature vectors are PF components.

[0054] Specifically, the LMD algorithm can adaptively decompose the target power quality interference signal into multiple PF components, where each PF component is a product combination of an envelope signal and a pure frequency modulation signal. The amplitude of each component is equal to the corresponding envelope signal, and its instantaneous frequency can be calculated through the pure frequency modulation signal. By multiplying the instantaneous frequency and amplitude obtained in the decomposition process, all PF components can be obtained, thereby generating a complete time-frequency spectrum of the target power quality interference signal. For a better understanding, considering the time series x(t) (target power quality interference signal), the main decomposition steps of the LMD algorithm are as follows:

[0055] Step 1: Determine all local extreme points n contained in the time series x(t) i, and calculate the extreme point n in turn i and n i+1 The average value m i ,in, Connect the calculated values ​​m with a broken line i And smooth it to get the local mean function m 11 (t);

[0056] Step 2: Set the adjacent extreme points n i and n i+1 The absolute difference between the two is divided by 2 to obtain the envelope estimate, where Similarly, use a broken line to connect the calculated values ​​a in sequence i And smooth it, we can get the envelope estimation function a 11 (t);

[0057] Step 3: Separate the local mean function m from the original time series x(t) 11 (t), where h 11 (t) = x(t) - m 11 (t), s 11 (t) = h 11 (t) / a 11 (t), in ideal conditions, s 11 (t) should be a pure frequency modulated signal, that is, its envelope estimation function a 12 (t) should be related to a 12 (t) = 1 matches if a 12 (t)≠1, then s 11 (t) is regarded as the original sequence, and the above cycle steps are repeated n times until s 1n (t) becomes a pure frequency modulated signal, satisfying condition a 1(n+1) (t) = 1, so the cycle process is as follows:

[0058] The general termination condition for the iteration is:

[0059] Step 4: The envelope signal is equal to the product of all envelope estimation functions obtained during the iterative decomposition process. The envelope signal represents the instantaneous amplitude function, and its mathematical expression is as follows:

[0060] Step 5: Compare the envelope signal a1(t) with the pure frequency modulated signal s 1n (t) to obtain the first PF component of LMD decomposition, where PF1(t) = a1(t)s 1n(t), the LMD algorithm decomposes the signal layer by layer from high frequency to low frequency, so the frequency of component PF1 is equal to the highest frequency of the sequence x(t). Theoretically, PF1 is a single-component AM / FM signal, and its instantaneous frequency and instantaneous amplitude are f1(t) and A1(t), respectively. A1(t) is equal to the envelope signal. a1(t)f1(t) can be obtained by 1n (t) Application equation get;

[0061] Step 6: Remove PF1(t) from the original sequence x(t) to obtain a new time series u1(t). Repeat steps 1 to 5 on u1(t). This process is repeated k times. k When it becomes a monotonic function, the above cycle ends. According to the above steps, k PF components can be obtained. Finally, the time series x(t) will be decomposed into multiple simple signals, including single components of k PF components and a monotonic residual component u k (t), where According to the above decomposition principle, it can be seen that the amplitude-frequency characteristics of the time series x(t) can be obtained by the linear combination of the amplitude and frequency of all PF components obtained by decomposition. Therefore, the LMD decomposition method retains the complete information of the original signal without causing any information loss, and thus has excellent integrity.

[0062] Step S2022, inputting the plurality of target feature vectors into a recognition model to obtain a plurality of target probabilities, wherein the recognition model is obtained by training an ELM model using a plurality of sets of training data, wherein each set of the training data at least includes: a plurality of feature vectors of the power quality interference signal, a plurality of predicted probabilities of the power quality interference signal acquired during a historical period, wherein one predicted probability corresponds to a predetermined type, the predicted probability is the probability that the type is the predetermined type corresponding to the predicted probability, and the target probability is the predicted probability of the target power quality interference signal;

[0063] In an optional solution, the above step S2022 may be implemented as follows:

[0064] Inputting a plurality of the target feature vectors into an input layer, wherein the input layer is composed of a plurality of input layer neurons, and one input layer neuron is used to receive one target feature vector;

[0065] Inputting the target feature vector received by the input layer into the hidden layer to obtain an output matrix, wherein the hidden layer is composed of a plurality of hidden layer neurons;

[0066] The output matrix is ​​input into the output layer to obtain a probability vector, the output layer is composed of a plurality of output layer neurons, and the probability vector includes all the target probabilities.

[0067] Specifically, Figure 3 As shown in the figure, the ELM model mainly consists of three parts: input layer, hidden layer and output layer. The input layer is composed of multiple input layer neurons a, and the input layer is used to receive the target feature vector x. The hidden layer is composed of multiple hidden layer neurons b. The output layer is composed of multiple output layer neurons c, and the output layer is used to output the probability vector y.

[0068] Specifically, the ELM model with M hidden layer neurons can be expressed by the following formula: k=1,2,N,where x k is the kth target feature vector, o k is the predetermined probability corresponding to the kth predetermined type, the predetermined probability is the probability that the target power quality interference signal is the corresponding predetermined type, g(.) represents the activation function, β i represents the weight vector between the i-th hidden layer neuron and the output layer neuron, w i represents the weight vector between the i-th hidden layer neuron and the input layer neuron, b i is the bias of the i-th hidden layer neuron, w i and x k The dot product between them is denoted as w i .x k , in order to minimize the output error, it can be expressed as: y k is the target probability corresponding to the kth type, that is, there exists b that satisfies the following formula i , β i and w i : k=1,2,N, which can be expressed in matrix form as follows: Hβ=Y, where H represents the output matrix of the hidden layer, Y represents the probability vector, and the output matrix H can be expressed as: In the ELM model, once the bias and weight w are randomly determined i , the output matrix of the hidden layer will become a fixed matrix. In this way, the problem can be transformed into solving the linear system of H-β=Y, without the need for a large number of parameter adjustments like traditional algorithms, which consumes a lot of time.

[0069] In an optional scheme, the number of the input layer neurons is the same as the number of the target probabilities, the number of the hidden layer neurons is twice the number of the target probabilities, the number of the output layer neurons is the same as the number of the target probabilities, and the bias of the hidden layer neurons is twice the number of the target probabilities.

[0070] In an optional solution, the above types include: normal voltage, voltage fluctuation, voltage instability, voltage interruption, instantaneous impulse, and transient oscillation.

[0071] Specifically, when the above types include: normal voltage, voltage fluctuation, voltage instability, voltage interruption, instantaneous impulse, and transient oscillation, the number of the above input layer neurons is 6, the number of the above hidden layer neurons is 12, the number of the above output layer neurons is 6, and the bias of the above hidden layer neurons is 12.

[0072] Step S2023: determining that the predetermined type corresponding to the maximum target probability is the type of the original power quality interference signal.

[0073] Through the above embodiment, first, the original power quality interference signal is denoised by the wavelet packet transform method to obtain the target power quality interference signal, and then the type of the original power quality interference signal is determined according to the target power quality interference signal, thereby solving the problem of inaccurate detection of the type of power quality interference signal caused by noise in the power quality interference signal in the prior art.

[0074] Example 1

[0075] The effectiveness of denoising the original power quality interference signal using the wavelet packet transform method is verified. Simulations are performed on Matlab software to verify the universality of power quality interference signals. Noise is added to the common single power quality interference signal and compared with commonly used signal denoising methods, including singular value decomposition denoising, wavelet adaptive threshold denoising and wavelet soft threshold denoising. The simulations are performed under various noise environments. Among these methods, the wavelet denoising method uses the db4 wavelet basis, which has good orthogonality and time-frequency support. The wavelet packet method uses a 6-level decomposition and selects the best wavelet packet tree based on the Shannon entropy principle. The fixed threshold method is used to select the threshold. The types of power quality interference signals considered in this application include voltage swell, voltage sag, voltage interruption, transient pulse and transient oscillation, as shown in Table 1. The formulas in Table 1 are used in MATLAB to generate these five signals, generating 1000 data samples respectively, and the data sets are divided into training and test sets in a ratio of 6:4.

[0076] Table 1

[0077]

[0078] When evaluating the denoising effects of different algorithms, two common evaluation indicators are usually used: signal-to-noise ratio (SNR) and mean square error (MSE). The signal-to-noise ratio reflects the relationship between the power quality interference signal and the noise, and to a certain extent indicates the smoothness of the power quality interference signal, while the mean square error reflects the delay characteristics of the signal.

[0079] The selected power quality interference signals used in this application mainly include voltage swell, voltage sag, voltage interruption, transient pulse and transient oscillation. The signal-to-noise ratio (SNR) and mean square error (MSE) of different denoising methods are as follows: Figure 4 and Figure 5 As shown, Figure 5 The rectangles with different gray levels represent the mean square errors of different types of power quality interference signals. Figure 4 and Figure 5 It can be found that compared with the traditional denoising method, the wavelet packet transform method of the present application shows better performance in terms of signal-to-noise ratio and mean square error indicators, and has the best denoising effect.

[0080] Example 2

[0081] In order to more intuitively demonstrate the feasibility and effectiveness of the proposed method, the recognition model in this application is compared with traditional classification models such as the SVM model and the BPNN model, as shown in Table 2. Table 2 lists the recognition accuracy of each method under different signal-to-noise ratio conditions and different types of power quality interference signals. The BPNN model has 6 neurons in the input layer, 12 neurons in the hidden layer, and 1 hidden layer, using sigmoid as the activation function. The learning rate is set to 0.01, the number of iterations is set to 200, the SVM model uses the RBF kernel, the penalty coefficient is 100, the γ value is 0.1, and the soft margin parameter is set to 0.01.

[0082] Table 2

[0083]

[0084]

[0085] As can be seen from Table 2 above, in the field of power quality disturbance classification, the ELM model has many advantages over traditional classification models such as the BPNN model and the SVM model. First, in terms of training speed and efficiency, the ELM model has obvious advantages. The ELM model is a fast single hidden layer feedforward neural network learning algorithm. In power quality disturbance identification, the ELM model can be used as a classifier to classify interference signals according to the extracted features. The ELM model has extremely fast training speed and good generalization ability. Compared with the BPNN model, the ELM model does not need to adjust parameters repeatedly, which greatly shortens the training time. In addition, the ELM model has high dimensional features and noise data. It has good robustness. The training process of the BPNN model involves multiple iterations and weight adjustments, which usually requires a lot of computing resources and time. On the contrary, the ELM model implements a single training process by randomly initializing input weights and biases and using analytical methods to solve output weights. This enables the ELM model to quickly adapt to new data and produce classification results, which is particularly evident when processing large-scale data sets. Although the training of the SVM model is also relatively fast, it usually requires the selection of appropriate kernel functions and parameters, while the training process of the ELM model is simpler and more efficient. Secondly, the ELM model performs well in generalization ability, which refers to the ability of the model to predict unseen data. Due to the random initialization and analytical solution methods used in the training process, the ELM model is able to learn the inherent patterns and features of the data, thereby achieving good generalization performance. In contrast, the BPNN model is prone to local optimal solutions, which may affect its generalization performance. The SVM model has good generalization ability, but its performance depends largely on the selection of kernel functions and parameter adjustment. In addition, from the perspective of parameter adjustment and model complexity, the ELM model still maintains its own advantages. The BPNN model requires adjustment of parameters such as weights, biases, and learning rates, which usually requires a lot of experiments and experience. The SVM model needs to select appropriate kernel functions and parameters, such as penalty coefficients and kernel function parameters, which also requires certain skills and experience. In contrast, the ELM model mainly involves adjusting the number of hidden layer nodes, which is a relatively simple parameter that has an intuitive impact on model performance. In addition, the structure of the ELM model is relatively simple, making it easy to understand and implement. In addition, the ELM model also has excellent stability and robustness. The random initialization and analytical solution methods used in the training process make the ELM model insensitive to the choice of initial parameters, thereby enhancing its stability. At the same time, the ELM model has a certain degree of robustness to noise and outliers, and can effectively resist data interference and fluctuations.

[0086] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0087] The embodiment of the present application also provides a power quality interference signal type detection device. It should be noted that the power quality interference signal type detection device of the embodiment of the present application can be used to execute the power quality interference signal type detection method provided by the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0088] The following is an introduction to the power quality interference signal type detection device provided in an embodiment of the present application.

[0089] Figure 6 is a structural block diagram of a device for detecting the type of power quality interference signal according to an embodiment of the present application. Figure 6 As shown, the device comprises:

[0090] A denoising unit 10 is used to perform denoising processing on the original power quality interference signal by using a wavelet packet transform method to obtain a target power quality interference signal;

[0091] In an optional solution, the denoising unit includes:

[0092] An acquisition module, used to acquire the sampling frequency of the original power quality interference signal and the base frequency of the power quality interference signal;

[0093] A determination module is used to determine the sampling frequency, the base frequency and Determine the number of divided frequency bands, where K is the number of divided frequency bands, f s is the above sampling frequency, f is the above base frequency;

[0094] The denoising module is used to denoise the original power quality interference signal based on the number of divided frequency bands and adopt the wavelet packet transform method to obtain the target power quality interference signal.

[0095] The determination unit 20 is used to determine the type of the original power quality interference signal according to the target power quality interference signal.

[0096] In an optional solution, the determining unit includes:

[0097] An extraction module, used for performing feature extraction on the target power quality interference signal to obtain a plurality of target feature vectors, wherein the target feature vectors are feature vectors of the target power quality interference signal;

[0098] In an optional solution, the extraction module is used to:

[0099] The target power quality interference signal is decomposed and processed by using the LMD algorithm to obtain a plurality of target feature vectors, where the target feature vectors are PF components.

[0100] An input module is used to input the multiple target feature vectors into the recognition model to obtain multiple target probabilities, wherein the recognition model is obtained by training the ELM model using multiple sets of training data, and each set of the training data at least includes: multiple feature vectors of the power quality interference signal, multiple predicted probabilities of the power quality interference signal acquired in a historical period, one of the predicted probabilities corresponds to a predetermined type, the predicted probability is the probability of the predetermined type corresponding to the predicted probability, and the target probability is the predicted probability of the target power quality interference signal;

[0101] In an optional solution, the input module includes:

[0102] A first input submodule is used to input a plurality of the target feature vectors into an input layer, wherein the input layer is composed of a plurality of input layer neurons, and one input layer neuron is used to receive one target feature vector;

[0103] A second input submodule is used to input the target feature vector received by the input layer into a hidden layer to obtain an output matrix, wherein the hidden layer is composed of a plurality of hidden layer neurons;

[0104] The third input submodule is used to input the above output matrix into the output layer to obtain a probability vector, the above output layer is composed of a plurality of output layer neurons, and the above probability vector contains all the above target probabilities.

[0105] In an optional scheme, the number of the input layer neurons is the same as the number of the target probabilities, the number of the hidden layer neurons is twice the number of the target probabilities, the number of the output layer neurons is the same as the number of the target probabilities, and the bias of the hidden layer neurons is twice the number of the target probabilities.

[0106] In an optional solution, the above types include: normal voltage, voltage fluctuation, voltage instability, voltage interruption, instantaneous impulse, and transient oscillation.

[0107] A determination module is used to determine that the predetermined type corresponding to the maximum target probability is the type of the original power quality interference signal.

[0108] Through the above embodiment, first, the original power quality interference signal is denoised by the wavelet packet transform method to obtain the target power quality interference signal, and then the type of the original power quality interference signal is determined according to the target power quality interference signal, thereby solving the problem of inaccurate detection of the type of power quality interference signal caused by noise in the power quality interference signal in the prior art.

[0109] The above-mentioned power quality interference signal type detection device includes a processor and a memory, and the above-mentioned denoising unit and determination unit are stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned modules are located in different processors in any combination.

[0110] The processor includes a kernel, and the kernel retrieves the corresponding program unit from the memory. One or more kernels can be set, and the problem of inaccurate detection of the type of power quality interference signal caused by noise in the power quality interference signal in the prior art is solved by adjusting the kernel parameters.

[0111] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0112] An embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the power quality interference signal type detection method.

[0113] Specifically, the power quality interference signal type detection method includes:

[0114] Step S201, using a wavelet packet transform method to perform denoising on the original power quality interference signal to obtain a target power quality interference signal;

[0115] Step S202: determining the type of the original power quality interference signal according to the target power quality interference signal.

[0116] An embodiment of the present invention provides a processor, and the processor is used to run a program, wherein the power quality interference signal type detection method is executed when the program is running.

[0117] Specifically, the power quality interference signal type detection method includes:

[0118] Step S201, using a wavelet packet transform method to perform denoising on the original power quality interference signal to obtain a target power quality interference signal;

[0119] Step S202: determining the type of the original power quality interference signal according to the target power quality interference signal.

[0120] An embodiment of the present invention provides a device, the device comprising a processor, a memory, and a program stored in the memory and executable on the processor, and when the processor executes the program, at least the following steps are implemented:

[0121] Step S201, using a wavelet packet transform method to perform denoising on the original power quality interference signal to obtain a target power quality interference signal;

[0122] Step S202: determining the type of the original power quality interference signal according to the target power quality interference signal.

[0123] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program for initializing at least the following method steps:

[0124] Step S201, using a wavelet packet transform method to perform denoising on the original power quality interference signal to obtain a target power quality interference signal;

[0125] Step S202: determining the type of the original power quality interference signal according to the target power quality interference signal.

[0126] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0127] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0128] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0129] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0130] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0131] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0132] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0133] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0134] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0135] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0136] 1) In the power quality interference signal type detection method of the present application, first, the original power quality interference signal is denoised using a wavelet packet transform method to obtain a target power quality interference signal, and then the type of the original power quality interference signal is determined based on the target power quality interference signal, thereby solving the problem of inaccurate power quality interference signal type detection caused by noise in the power quality interference signal in the prior art.

[0137] 2) In the power quality interference signal type detection device of the present application, first, the original power quality interference signal is denoised by a wavelet packet transform method to obtain a target power quality interference signal, and then the type of the original power quality interference signal is determined according to the target power quality interference signal, thereby solving the problem of inaccurate detection of the power quality interference signal type caused by noise in the power quality interference signal in the prior art.

[0138] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for detecting the type of power quality interference signal, characterized in that: The method comprises: The wavelet packet transform method is used to denoise the original power quality interference signal to obtain the target power quality interference signal; The type of the original power quality interference signal is determined according to the target power quality interference signal.

2. The method according to claim 1, characterized in that The wavelet packet transform method is used to denoise the original power quality interference signal to obtain the target power quality interference signal, including: Acquire the sampling frequency of the original power quality interference signal and the fundamental frequency of the power quality interference signal; According to the sampling frequency, the base frequency and Determine the number of divided frequency bands, where K is the number of divided frequency bands, f s is the sampling frequency, and f is the base frequency; Based on the number of divided frequency bands, the wavelet packet transform method is used to perform denoising processing on the original power quality interference signal to obtain the target power quality interference signal.

3. The method according to claim 1, characterized in that: Determining the type of the original power quality interference signal according to the target power quality interference signal includes: Extracting features of the target power quality interference signal to obtain a plurality of target feature vectors, wherein the target feature vectors are feature vectors of the target power quality interference signal; Inputting a plurality of the target feature vectors into a recognition model to obtain a plurality of target probabilities, wherein the recognition model is obtained by training an ELM model using a plurality of sets of training data, wherein each set of the training data at least includes: a plurality of the feature vectors of the power quality interference signal, a plurality of predicted probabilities of the power quality interference signal acquired during a historical period, wherein one predicted probability corresponds to a predetermined type, the predicted probability is the probability of the predetermined type corresponding to the predicted probability, and the target probability is the predicted probability of the target power quality interference signal; The predetermined type corresponding to the maximum target probability is determined to be the type of the original power quality interference signal.

4. The method according to claim 3, characterized in that Feature extraction is performed on the target power quality interference signal to obtain multiple target feature vectors, including: The target power quality interference signal is decomposed and processed by using the LMD algorithm to obtain a plurality of target feature vectors, where the target feature vectors are PF components.

5. The method according to claim 3, characterized in that: Inputting a plurality of the target feature vectors into a recognition model, a plurality of target probabilities are obtained, including: Inputting a plurality of the target feature vectors into an input layer, wherein the input layer is composed of a plurality of input layer neurons, and one input layer neuron is used to receive one target feature vector; Inputting the target feature vector received by the input layer into a hidden layer to obtain an output matrix, wherein the hidden layer is composed of a plurality of hidden layer neurons; The output matrix is ​​input into an output layer to obtain a probability vector, wherein the output layer is composed of a plurality of output layer neurons, and the probability vector contains all the target probabilities.

6. The method according to claim 5, characterized in that The number of neurons in the input layer is the same as the number of target probabilities, the number of neurons in the hidden layer is twice the number of target probabilities, the number of neurons in the output layer is the same as the number of target probabilities, and the bias of the neurons in the hidden layer is twice the number of target probabilities.

7. The method according to any one of claims 1 to 6, characterized in that: The types include: normal voltage, voltage fluctuation, voltage instability, voltage interruption, instantaneous impulse, and transient oscillation.

8. A device for detecting the type of power quality interference signal, characterized in that: The device comprises: A denoising unit, used for denoising the original power quality interference signal by using a wavelet packet transform method to obtain a target power quality interference signal; A determination unit is used to determine the type of the original power quality interference signal according to the target power quality interference signal.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the power quality interference signal type detection method according to any one of claims 1 to 7.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the method for detecting the type of power quality interference signal described in any one of claims 1 to 7 is implemented.