Composite power quality disturbance parameter identification method based on ISVMD-HT

Abstract

The present invention proposes a composite power quality disturbance parameter identification method based on ISVMD-HT. First, the continuous variational modal decomposition algorithm is improved by optimal matching extension to overcome the endpoint effect and modal aliasing defects in signal decomposition to a large extent; then the composite disturbance signal is decomposed by using a new continuous variational modal decomposition, and the decomposed modal components are Hilbert transformed to obtain the instantaneous amplitude spectrum of the disturbance signal; finally, the instantaneous amplitude spectrum analysis is combined to realize the parameter identification of the composite power quality disturbance. The present invention helps to realize a more accurate identification of the increasingly complex composite power quality disturbance parameters in the new power electronic power system.

Description

Technical Field

[0001] The invention belongs to the technical field of parameter identification of composite power quality disturbance signals, and in particular relates to a composite power quality disturbance parameter identification method based on ISVMD-HT. Background Art

[0002] The identification of composite power quality disturbance parameters mainly involves identifying the parameters of each single power quality disturbance signal that constitutes the composite disturbance. The signal parameters mainly include the start and end time of the disturbance, the amplitude and the frequency change. Accurately identifying the signal parameters of the composite power quality disturbance is extremely important for further understanding the degree of power quality disturbance and targeted treatment of power quality disturbance.

[0003] At present, the methods for identifying power quality disturbance parameters can be mainly divided into parametric methods and non-parametric methods. Commonly used parametric methods include Multiple Signal Classification (MUSIC), Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) and Kalman Filter (KF). The MUSIC algorithm is often used for parameter detection of steady-state disturbances such as harmonics and interharmonics. In view of the problems of fixed window size and poor resolution of the traditional sliding window ESPRIT algorithm, researchers proposed an adaptive sliding window ESPRIT algorithm. The time-frequency characteristics of the algorithm have been greatly improved, but it is difficult to apply to increasingly complex multiple composite power quality disturbances. Commonly used non-parametric methods include Fourier transform, HHT transform, singular value decomposition, wavelet transform, S transform and empirical wavelet method. HHT has good applicability to composite power quality disturbance signals, but the two inherent defects of the end effect and modal aliasing of its algorithm components EMD are still unresolved. Although EEMD can reduce the impact, it still has certain limitations and cannot be applied to the detection of disturbance signals in actual complex situations. Some researchers have proposed a new method for power quality disturbance detection that combines improved variational mode decomposition (VMD) with Hilberet transform (HT). This paper first determines the pre-decomposition scale K of VMD transform by Fourier transform, and then uses VMD-HT transform to determine the instantaneous amplitude spectrum, thereby realizing the amplitude-frequency parameter identification of the disturbance signal. However, in practice, some disturbance components with similar frequencies are difficult to separate by Fourier transform. Singular value decomposition does not require pre-adjustment of parameters and has high detection accuracy for transient disturbance detection, but it has certain limitations for disturbance detection occurring at zero crossings. Some scholars extract signal features through sliding window singular value decomposition and obtain positioning information through gradients. While solving the zero crossing problem, it has the characteristics of accurate positioning and strong anti-interference ability. Wavelet packet transform is very prone to spectrum leakage in the process of decomposing composite disturbances, which leads to large deviations in disturbance parameter detection. The S transform can process non-stationary signals well, but for the increasingly complex types of power quality disturbances, this method has a large amount of calculation and low practicality. The empirical wavelet method can adaptively decompose most non-stationary signals, and combined with the HT transform, it can achieve high-precision parameter identification of composite power quality disturbances. The power quality disturbance detection algorithm based on mathematical morphology theory has achieved certain results. Although its operation speed is fast, the uncertainty of its structural element selection brings certain limitations to the algorithm, and it cannot be fully applied to more complex and diverse composite power quality disturbances.Therefore, there is an urgent need to seek an efficient and high-precision composite disturbance signal parameter identification method, which can not only effectively decompose the increasingly complex multiple composite disturbance signals under the current power electronics new power system, but also identify the signal parameters of each disturbance component in the composite disturbance with high precision. Summary of the invention

[0004] Considering the defects and shortcomings of the prior art, the present invention proposes a composite power quality disturbance parameter identification method based on ISVMD-HT, which is used to solve the problem that in a new power system in which a large number of power electronic components are incorporated, the types of power quality disturbances become more and more complicated and the parameters are difficult to accurately identify.

[0005] It is mainly composed of the best matching extension, improved continuous variational modal decomposition and Hilbert transform technology. First, the continuous variational modal decomposition algorithm is improved by the best matching extension, so that it can overcome the endpoint effect and modal aliasing defects in signal decomposition to the greatest extent; then the composite disturbance signal is decomposed by the new continuous variational modal decomposition, and the decomposed modal components are Hilbert transformed to obtain the instantaneous amplitude spectrum of the disturbance signal; finally, the parameter identification of the composite power quality disturbance is realized by combining the instantaneous amplitude spectrum analysis. The present invention helps to realize more accurate identification of the increasingly complex composite power quality disturbance parameters in the new power electronic power system. Not only can the composite disturbance be decomposed efficiently and accurately without the need to determine the number of disturbances in the composite disturbance in advance, but the combination of Hilbert transform instantaneous amplitude spectrum analysis can also realize high-precision identification of the parameters of each disturbance component in the composite disturbance, which is of great significance to the safe and stable operation and regulation of the power grid.

[0006] The present invention specifically adopts the following technical solutions:

[0007] A composite power quality disturbance parameter identification method based on ISVMD-HT is characterized by: improving the continuous variational modal decomposition algorithm through optimal matching extension, performing Hilbert transform on the modal components decomposed by the improved continuous variational modal decomposition algorithm respectively, and obtaining the Hilbert instantaneous amplitude spectrum corresponding to each modal component; locating the start and end time of each disturbance component by analyzing the instantaneous frequency spectrum, and tracking and determining the amplitude change of each disturbance component by using its instantaneous amplitude spectrum; finally, obtaining high-precision parameter identification of each disturbance component in the composite disturbance by processing and analyzing the instantaneous spectrum.

[0008] Furthermore, the continuous variational mode decomposition algorithm is improved by best matching extension, which specifically includes the following steps:

[0009] Step S1: Sample the original signal x(t) and obtain the extreme point sequence n of the signal i ;

[0010] Step S2: Select a characteristic waveform, and take the waveform including the left endpoint x(1) and the extreme points x(n1), x(n2), and x(n3) as the characteristic waveform W, with a length of L, and the extreme point x(n3) is the right endpoint of the characteristic waveform;

[0011] Step S3: intercept the matching waveform and in the subsequent signal, use x(n 2k+1 ) as the right endpoint, take the waveform of length L to the left as the matching waveform W k (k=2,3,…);

[0012] Step S4: Taking curvature matching as the key indicator for optimal matching extension, the curvature near the sampling point in the band is calculated as the characteristic curvature C according to the following formula:

[0013]

[0014] In the formula, i is the serial number of the sampling point in the band; is the first-order derivative of the ith sampling point, is the second-order derivative of the i-th sampling point;

[0015] Step S5: Select the correlation coefficient ρ, curvature matching error ε and absolute error σ as indicators for waveform matching screening; use formula (1) to calculate the curvature C near the sampling point in the k waveform segments k (i) Calculate the correlation coefficient, matching error and waveform absolute error between the characteristic band curvature and the matching band curvature according to the following formulas;

[0016]

[0017]

[0018]

[0019] Where Cov(C,C k ) represents the covariance between the curvature of the characteristic band and the curvature of the kth matching band; D(C) is the variance of the curvature of the characteristic waveform, D(C k ) is the variance of the curvature of the kth matching waveform; σ k0 Represents the absolute error between the feature band and the kth matching band; N0 is the number of sampling points in the feature band, and N is the total number of sampling points;

[0020] Step S6: Using the matching degree P k0 The best matching waveform is selected based on the comparison index of the best matching extension. When the absolute error of the kth waveform segment satisfies σ<α*L and the matching degree P k0 Maximum, in W kis the best matching waveform segment; the previous point x(i-1) of its left endpoint x(i) is used as the right endpoint of the extended waveform, and a waveform with a length of u is intercepted to the left as the extended waveform; the size of u is determined according to the situation of the disturbance signal; the extended waveform is moved left to before x(1) to complete the left endpoint extension of the signal; and the right endpoint extension of the signal is completed according to the same rule:

[0021]

[0022] Where α is a constant, which is adjusted according to the actual signal; P k0 is the matching index between the kth matching waveform segment and the characteristic waveform segment;

[0023] If the best matching waveform segment is not found, it indicates that the intrinsic curvature law is not suitable for the curvature change trend of the characteristic waveform selected by the boundary, that is, the length of the perturbed part or the non-perturbed part of the waveform at the end point of the waveform is too short, and no matching waveform segment can be found; at this time, the extreme point closest to the end point is selected as the mirror symmetry axis of the mirror extension to complete the extension of the signal at the end point;

[0024] Step S7: Perform continuous variational mode decomposition on the extended signal and initialize the penalty factor α and error ε; assuming that the input signal is decomposed into L eigenmodes, then:

[0025] f(t)=u L (t)+f r (t) (6)

[0026]

[0027] In the formula, u L (t) is the Lth eigenmode function, f r (t) represents the residual component, f r (t) represents the sum of the first L-1 modes and the undecomposed partial signal f u (t);

[0028] Step S8: Construct the constraints of continuous variational mode decomposition (SVMD): J1, J2 and J3:

[0029] (1) The frequency band of each modal function should converge to ω L ,ω L is the center frequency of the Lth mode function;

[0030]

[0031] (2) Residual component f r (t) is minimized when the Lth mode function is meaningful:

[0032]

[0033]

[0034] In the formula, is the filter operator, in order to minimize f r (t) and u L (t) modal aliasing, filtered by the filter operator f r (t) Energy should be minimized;

[0035] (3) Lth modal function u L (t) The energy near the center frequency of other modes should be minimized:

[0036]

[0037] (4) All modal functions and f u The reconstructed signal obtained by summing (t) should be equal to the extended signal:

[0038]

[0039] Step S9: Add a quadratic penalty factor and construct a Lagrangian constraint problem. By solving the Lagrangian optimal solution, the SVMD decomposition of the extended signal is realized:

[0040]

[0041] Step S10: according to step S6, the waveform length u is extended, and the extended parts are respectively intercepted for the decomposed modal components to obtain an improved continuous variational modal decomposition ISVMD of the signal.

[0042] Furthermore, the modal components decomposed by the improved continuous variational modal decomposition ISVMD are subjected to Hilbert transform to obtain the instantaneous amplitude spectrum of each modal component; the start and end time of the disturbance of each disturbance component can be located by analyzing the instantaneous frequency spectrum, and the amplitude change of each disturbance component can be tracked and determined using its instantaneous amplitude spectrum; thereby, accurate identification of the parameters of each disturbance component of the composite power quality disturbance is achieved.

[0043] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the composite power quality disturbance parameter identification method based on ISVMD-HT as described above is implemented.

[0044] A computer-readable storage medium stores a computer program thereon, characterized in that when the program is executed by a processor, the composite power quality disturbance parameter identification method based on ISVMD-HT as described above is implemented.

[0045] Compared with the prior art, the present invention and its preferred embodiments have at least the following outstanding advantages:

[0046] 1. The SVMD algorithm is improved by the best matching extension algorithm, which greatly reduces the decomposition error caused by the endpoint effect of VMD and traditional SVMD algorithms. Moreover, ISVMD converts the K-dimensional optimization problem in VMD decomposition into K one-dimensional optimization problems by continuously and one by one solving each modal function. This not only greatly reduces the computational complexity of the decomposition algorithm, but also does not require the pre-knowledge of the modal decomposition number K, which increases the adaptability of the algorithm. The ISVMD algorithm can effectively decompose more complex composite power quality disturbance signals in power electronic power systems.

[0047] 2. Based on ISVMD-HT, composite power quality parameter identification can perform Hilbert transform on the decomposed effective components to obtain their corresponding instantaneous amplitude spectrum. By analyzing the instantaneous frequency spectrum of each modal component, the start and end time of each disturbance component can be accurately located; by analyzing the instantaneous amplitude spectrum of each modal component, the amplitude change of each disturbance component can be accurately tracked. It can be seen that the instantaneous amplitude spectrum extracted by ISVMD-HT can identify the signal parameters of composite power quality disturbance with high accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 The figure is a schematic diagram of the overall flow of the method according to the embodiment of the present invention.

[0049] Figure 2 This is a schematic diagram of an example of best matching extension according to an embodiment of the present invention.

[0050] Figure 3 Schematic diagram of parameter identification results of composite power quality disturbance test signal ISVMD-HT (instantaneous frequency spectrum of composite disturbance signal ISVMD-HT) according to an embodiment of the present invention.

[0051] Figure 4 Schematic diagram of parameter identification results of composite power quality disturbance test signal ISVMD-HT (instantaneous amplitude spectrum of composite disturbance signal ISVMD-HT) according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the features and advantages of this patent more obvious and easy to understand, the following embodiments are specifically described in detail as follows:

[0053] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used in this specification have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0054] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0055] like Figure 1-Figure 4 As shown, the composite power quality disturbance parameter identification method based on ISVMD-HT provided in this embodiment can be described as the following steps:

[0056] Step S1: Figure 2 The composite power quality disturbance test signal x(t) is shown, and the extreme point sequence n of the signal is defined as i (i=1,2,...);

[0057] Step S2: Select a characteristic waveform, and take the waveform containing x(1), x(n1), x(n2), and x(n3) as the characteristic waveform W, with a length of L, where x(1) is the left endpoint of the waveform and x(n3) is the right endpoint of the waveform.

[0058] Step S3: intercept the matching waveform, and in the subsequent signal, use x(n 2k+1 ) as the right endpoint, take a waveform of length L to the left as the matching waveform W k (k=2,3,…).

[0059] Step S4: Taking curvature matching as the key indicator for optimal matching extension, the curvature near the sampling point in the band is calculated according to the following formula as the characteristic curvature C:

[0060]

[0061] Where i is the serial number of the sampling point in the band; y i is the first-order derivative of the i-th sampling point, y i is the second-order derivative of the ith sampling point.

[0062] Step S5: Select the correlation coefficient ρ, curvature matching error ε and absolute error σ as indicators for waveform matching screening. Use formula (1) to calculate the curvature C near the sampling point in the k waveform segments: k (i). Calculate the correlation coefficient, matching error and waveform absolute error between the characteristic band curvature and the matching band curvature according to the following formulas;

[0063]

[0064]

[0065]

[0066] Where Cov(C,C k ) represents the covariance between the curvature of the characteristic band and the curvature of the kth matching band; D(C) is the variance of the curvature of the characteristic waveform, D(C k ) is the variance of the curvature of the kth matching waveform; σ k0 It represents the absolute error between the feature band and the kth matching band; N0 is the number of sampling points in the feature band, and N is the total number of sampling points.

[0067] Step S6: Using the matching degree P k0 The best matching waveform is selected based on the comparison index of the best matching extension. When the absolute error of the kth waveform segment satisfies σ<α*L and the matching degree P k0 Maximum, in W k is the best matching waveform segment. Take the previous point x(i-1) of its left endpoint x(i) as the right endpoint of the extended waveform, and intercept the waveform with a length of u to the left as the extended waveform. The size of u is determined according to the condition of the disturbance signal. Move the extended waveform to the left before x(1) to complete the left endpoint extension of the signal. Use a similar method to complete the right endpoint extension of the signal.

[0068]

[0069] Where α is a constant and needs to be adjusted according to the actual signal; P k0 is the matching index between the kth matching waveform segment and the characteristic waveform segment;

[0070] If the best matching waveform segment is not found, it means that the intrinsic curvature law is not suitable for the curvature change trend of the characteristic waveform selected by the boundary, that is, the length of the perturbed part or the non-perturbed part of the waveform at the end point of the waveform is too short, and no matching waveform segment can be found. The extreme point closest to the end point is selected as the mirror symmetry axis of the mirror extension to complete the extension of the signal at the end point.

[0071] Step S7: Perform continuous variational mode decomposition on the extended signal and initialize the penalty factor α and error ε. Assuming that the input signal is decomposed into L eigenmodes, we have:

[0072] f(t)=u L (t)+f r (t) (6)

[0073]

[0074] In the formula, u L (t) is the Lth eigenmode function, f r (t) represents the residual component, f r(t) represents the sum of the first L-1 modes and the undecomposed partial signal f u (t).

[0075] Step S8: Construct the SVMD constraint:

[0076] (1) The frequency band of each modal function should converge to ω L ,ω L is the center frequency of the Lth mode function;

[0077]

[0078] (2) Residual component f r (t) minimized when the Lth mode function is meaningful;

[0079]

[0080]

[0081] In the formula, is the filter operator, in order to minimize f r (t) and u L (t) modal aliasing, filtered by the filter operator f r (t) Energy should be minimized.

[0082] (3) Lth modal function u L (t) The energy around the center frequencies of other modes should be minimized:

[0083]

[0084] (4) All modal functions and f u The reconstructed signal obtained by summing (t) should be equal to the extended signal:

[0085]

[0086] Step S9: In order to make the decomposition have good convergence in a noisy environment, a quadratic penalty factor is added and a Lagrangian constraint problem is constructed. By solving the Lagrangian optimal solution, the SVMD decomposition of the extended signal is realized.

[0087]

[0088] Step S10: According to step S6, the waveform length u is extended, and the extended parts of the decomposed modal components are respectively intercepted, so as to finally obtain the improved successive variational modal decomposition (ISVMD) of the signal.

[0089] Step S11: Figure 3 , Figure 4 As shown, this embodiment decomposes the test signal through the ISVMD algorithm, and the composite disturbance signal is decomposed into 4 modal components, corresponding to the fundamental component, the third harmonic, the fifth harmonic and the seventh harmonic component. Then, each modal component after ISVMD decomposition is subjected to Hilbert transform to obtain the instantaneous amplitude spectrum of each modal component. By analyzing the instantaneous frequency spectrum, the start and end time of each disturbance component disturbance can be located, and the amplitude change of each disturbance component can be tracked and determined using its instantaneous amplitude spectrum. In this way, the accurate identification of the parameters of each disturbance component of the composite power quality disturbance can be achieved.

[0090] In summary, the present invention helps to achieve more accurate identification of increasingly complex composite power quality disturbance parameters in new power electronic power systems. It can not only efficiently and accurately decompose composite disturbances without the need to determine the number of disturbances in the composite disturbances in advance, but also combine Hilbert transform instantaneous amplitude spectrum analysis to ultimately achieve high-precision identification of composite disturbance signal parameters, thereby improving the stable operation capability of the power grid.

[0091] The above-mentioned program design scheme involving the algorithm provided in this embodiment can be stored in a computer-readable storage medium in a coded form, and implemented in the form of a computer program, and the basic parameter information required for the calculation is input through computer hardware, and the calculation result is output.

[0092] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0093] The present invention is described with reference to flowcharts of methods, devices (apparatus), and computer program products according to embodiments of the present invention. It should be understood that each process in the flowchart, as well as the combination of processes in the flowchart, 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 produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process in the flowchart. Figure 1 A device that specifies functions in a process or multiple processes.

[0094] 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 functions specified in a flowchart.

[0095] 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 The steps of a specified function in a process or multiple processes.

[0096] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any technician familiar with the profession may use the above disclosed technical content to change or modify it into an equivalent embodiment with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention still belongs to the protection scope of the technical solution of the present invention.

[0097] This patent is not limited to the above-mentioned optimal implementation mode. Anyone can derive other various forms of composite power quality disturbance parameter identification methods based on ISVMD-HT under the inspiration of this patent. All equal changes and modifications made according to the scope of the patent application of the present invention should be covered by this patent.

Claims

1. A composite power quality disturbance parameter identification method based on ISVMD-HT, characterized by: The continuous variational modal decomposition algorithm is improved through optimal matching extension. The modal components decomposed by the improved continuous variational modal decomposition algorithm are subjected to Hilbert transformation respectively to obtain the Hilbert instantaneous amplitude spectrum corresponding to each modal component. The start and end time of each disturbance component is located by analyzing the instantaneous frequency spectrum, and the amplitude change of each disturbance component can be tracked and determined by using its instantaneous amplitude spectrum. Finally, the high-precision parameter identification of each disturbance component in the composite disturbance is obtained by analyzing the instantaneous spectrum. The improvement of the continuous variational mode decomposition algorithm by best matching extension specifically includes the following steps: Step S1: Sample the original signal x(t) and obtain the extreme point sequence n of the signal i ; Step S2: Select a characteristic waveform, and take the waveform including the left endpoint x(1) and the extreme points x(n1), x(n2), and x(n3) as the characteristic waveform W, with a length of Lo, and the extreme point x(n3) is the right endpoint of the characteristic waveform; Step S3: intercept the matching waveform and in the subsequent signal, use x(n 2k+1 ) as the right endpoint, and take the waveform of length Lo to the left as the matching waveform W k (k=2,3,…); Step S4: Taking curvature matching as the key indicator for optimal matching extension, the curvature near the sampling point in the band is calculated as the characteristic curvature C according to the following formula: In the formula, i is the serial number of the sampling point in the band; is the first-order derivative of the ith sampling point, is the second-order derivative of the i-th sampling point; Step S5: Select the correlation coefficient ρ, the curvature matching error ε and the absolute error σ as indicators for waveform matching screening; Use formula (1) to calculate the curvature C near the sampling point in the k waveform segment k (i) Calculate the correlation coefficient, matching error and waveform absolute error between the characteristic band curvature and the matching band curvature according to the following formulas; Where Cov(C,C k ) represents the covariance between the curvature of the characteristic band and the curvature of the kth matching band; D(C) is the variance of the curvature of the characteristic waveform, D(C k ) is the variance of the curvature of the kth matching waveform; σ k0 Represents the absolute error between the feature band and the kth matching band; N0 is the number of sampling points in the feature band; Step S6: Using the matching degree P k0 The best matching waveform is selected based on the comparison index of the best matching extension. When the absolute error of the kth waveform segment satisfies σ<γ*Lo and the matching degree P k0 Maximum, in W k is the best matching waveform segment; the previous point x(i-1) of its left endpoint x(i) is used as the right endpoint of the extended waveform, and a waveform with a length of u is intercepted to the left as the extended waveform; the size of u is determined according to the situation of the disturbance signal; the extended waveform is moved left to before x(1) to complete the left endpoint extension of the signal; and the right endpoint extension of the signal is completed according to the same rule: Where γ is a constant, which is adjusted according to the actual signal; P k0 is the matching index between the kth matching waveform segment and the characteristic waveform segment; If the best matching waveform segment is not found, it indicates that the intrinsic curvature law is not suitable for the curvature change trend of the characteristic waveform selected by the boundary, that is, the length of the perturbed part or the non-perturbed part of the waveform at the end point of the waveform is too short, and no matching waveform segment can be found; at this time, the extreme point closest to the end point is selected as the mirror symmetry axis of the mirror extension to complete the extension of the signal at the end point; Step S7: Perform continuous variational mode decomposition on the extended signal and initialize the penalty factor α and error ε; assuming that the input signal is decomposed into L eigenmodes, then: f(t)=u L (t)+f r (t) (6) In the formula, u L (t) is the Lth eigenmode function, f r (t) represents the residual component, f r (t) is represented by the sum of the first L-1 modes and the undecomposed partial signal f u (t) and; Step S8: Construct the constraints of continuous variational mode decomposition (SVMD): J1, J2 and J3: (1) The frequency band of each modal function should converge to ω L ,ω L is the center frequency of the Lth mode function; (2) Residual component f r (t) is minimized when the Lth mode function is meaningful: In the formula, is the filter operator, in order to minimize f r (t) and u L (t) modal aliasing, filtered by the filter operator f r (t) Energy should be minimized; (3) Lth modal function u L (t) The energy near the center frequency of other modes should be minimized: (4) All modal functions and f u The reconstructed signal obtained by summing (t) should be equal to the extended signal: Step S9: Add a quadratic penalty factor and construct a Lagrangian constraint problem. By solving the Lagrangian optimal solution, the SVMD decomposition of the extended signal is realized: Step S10: according to step S6, the waveform length u is extended, and the extended parts are respectively intercepted for the decomposed modal components to obtain an improved continuous variational modal decomposition ISVMD of the signal.

2. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the composite power quality disturbance parameter identification method based on ISVMD-HT as claimed in claim 1 is implemented.

3. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the composite power quality disturbance parameter identification method based on ISVMD-HT as claimed in claim 1 is implemented.

Images