A method and system for parameter identification of a mixed oscillatory signal

Abstract

The application discloses a kind of mixed oscillation signal parameter identification method and system, it is related to hydroelectric parameter identification technical field.The mixed oscillation signal parameter identification method includes: obtaining mixed oscillation signal;Using second-order synchronous extraction wavelet transform method is decomposed to the mixed shock oscillation signal, and the decomposition signal is obtained;According to loss precision, whether the decomposition signal is completely decomposed is determined;The decomposition signal of complete decomposition is carried out empirical envelope demodulation, and the oscillation dominant mode of the decomposition signal of complete decomposition is obtained;The oscillation dominant mode includes low-frequency oscillation and ultra-low frequency oscillation.The application can accurately analyze the characteristic parameter of to-be-measured signal, and then accurately distinguish the oscillation dominant mode of signal.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydroelectric parameter identification, in particular to a hybrid oscillation signal parameter identification method and system. BACKGROUND

[0002] For a long time, in the asynchronous interconnection system with high penetration of hydroelectric generating units, the rotational inertia is only 1 / 6-1 / 5 of the synchronous time, the power grid anti-interference ability is reduced, the frequency oscillation risk is increased, and the hybrid oscillation phenomenon of low-frequency oscillation and ultra-low-frequency oscillation is prone to occur, which endangers the safe and stable operation of the power system. Therefore, it is crucial to effectively extract the hybrid oscillation mode and accurately obtain the parameters. However, the current time-frequency analysis method generally has low time-frequency resolution and poor noise immunity, and cannot accurately analyze the characteristic parameters of the measured signal. SUMMARY

[0003] The purpose of the present application is to provide a hybrid oscillation signal parameter identification method and system, which can accurately analyze the characteristic parameters of the measured signal, and accurately identify the dominant oscillation mode of the signal.

[0004] To achieve the above purpose, the present application provides the following scheme:

[0005] A hybrid oscillation signal parameter identification method, comprising:

[0006] obtaining a hybrid oscillation signal;

[0007] using a second-order synchronous extraction wavelet transform method to decompose the hybrid oscillation signal to obtain a decomposition signal;

[0008] determining whether the decomposition signal is completely decomposed according to the loss precision;

[0009] performing empirical envelope demodulation on the completely decomposed decomposition signal to obtain the dominant oscillation mode of the completely decomposed decomposition signal; the dominant oscillation mode includes low-frequency oscillation and ultra-low-frequency oscillation.

[0010] Optionally, the determination of whether the decomposition signal is completely decomposed according to the loss precision specifically comprises:

[0011] reconstructing a plurality of the decomposition signals to obtain a reconstructed signal;

[0012] judging whether the loss precision of the reconstructed signal is less than a preset threshold value;

[0013] If yes, the decomposition signal is completely decomposed, and the completely decomposed decomposition signal is output.

[0014] Optionally, the empirical envelope demodulation on the completely decomposed decomposition signal to obtain the dominant oscillation mode of the completely decomposed decomposition signal specifically comprises:

[0015] empirical envelope demodulation is performed on the completely decomposed decomposition signal to obtain mode information; the mode information at least includes amplitude, frequency, attenuation factor and damping ratio;

[0016] a dominant mode of oscillation of the completely decomposed decomposition signal is determined according to the mode information.

[0017] Optionally, the empirical envelope demodulation performed on the completely decomposed decomposition signal to obtain mode information specifically includes:

[0018] maximum value operation is performed on the completely decomposed decomposition signal to obtain maximum value and corresponding time;

[0019] cubic spline function is adopted to fit the maximum value and the corresponding time to obtain an empirical envelope function;

[0020] standardization processing is performed on the completely decomposed decomposition signal according to the empirical envelope function to obtain a pure frequency modulation function;

[0021] mode information is determined according to the completely decomposed decomposition signal and the pure frequency modulation function.

[0022] The application further discloses a hybrid oscillation signal parameter identification system, comprising:

[0023] a data acquisition module configured to acquire a hybrid oscillation signal;

[0024] a signal decomposition module configured to decompose the hybrid oscillation signal by adopting a second-order synchronous extraction wavelet transform method to obtain a decomposition signal;

[0025] a signal judgment module configured to determine whether the decomposition signal is completely decomposed according to loss precision;

[0026] a mode determination module configured to perform empirical envelope demodulation on the completely decomposed decomposition signal to obtain a dominant mode of oscillation of the completely decomposed decomposition signal; the dominant mode of oscillation includes low-frequency oscillation and ultra-low-frequency oscillation.

[0027] Optionally, the signal judgment module includes:

[0028] a signal reconstruction unit configured to reconstruct a plurality of the decomposition signals to obtain a reconstructed signal;

[0029] a signal judgment unit configured to judge whether loss precision of the reconstructed signal is less than a preset threshold value;

[0030] a signal output unit configured to output the completely decomposed decomposition signal when the loss precision of the reconstructed signal is less than the preset threshold value, and the decomposition signal is completely decomposed.

[0031] Optionally, the mode determination module comprises:

[0032] a mode information determination unit configured to perform empirical envelope demodulation on the completely decomposed decomposition signal to obtain mode information, wherein the mode information comprises at least amplitude, frequency, attenuation factor and damping ratio;

[0033] a mode determination unit configured to determine the dominant oscillation mode of the completely decomposed decomposition signal according to the mode information.

[0034] Optionally, the mode information determination unit comprises:

[0035] a data operator unit configured to perform maximum value operation on the completely decomposed decomposition signal to obtain maximum values and corresponding time instants;

[0036] a data fitting subunit configured to fit the maximum values and the corresponding time instants by using a cubic spline function to obtain an empirical envelope function;

[0037] a standardization processing subunit configured to perform standardization processing on the completely decomposed decomposition signal according to the empirical envelope function to obtain a pure frequency modulation function;

[0038] a mode information determination subunit configured to determine mode information according to the completely decomposed decomposition signal and the pure frequency modulation function.

[0039] According to the embodiments of the present application, the following technical effects are provided:

[0040] The application discloses a mixed oscillation signal parameter identification method and system. The application first adopts a second-order synchronous extraction wavelet transform method to decompose a mixed oscillation signal to obtain a decomposition signal, and determines whether the decomposition signal is completely decomposed according to loss precision. Then, empirical envelope demodulation is performed on the completely decomposed decomposition signal to obtain a corresponding dominant oscillation mode of the decomposition signal, wherein the dominant oscillation mode comprises low-frequency oscillation and ultra-low-frequency oscillation. The second-order synchronous extraction wavelet transform method is adopted to remove noise components irrelevant to real frequency, avoid noise interference and mode aliasing, improve time-frequency resolution and noise immunity, and the empirical envelope demodulation is adopted to obtain the corresponding dominant oscillation mode of the decomposition signal to complete parameter identification, so that the complex integral transform process is omitted, the end effect is avoided, and the characteristic parameters of the to-be-measured signal can be accurately analyzed. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described below only show some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0042] Figure 1 The method flow chart of the mixed oscillation signal parameter identification method of the present application;

[0043] Figure 2 The logic diagram of the mixed oscillation signal parameter identification method of the present application;

[0044] Figure 3 The simulation signal waveform diagram in the first experimental example of the present application;

[0045] Figure 4 The first decomposition signal diagram obtained by the SSEWT algorithm in the first experimental example of the present application;

[0046] Figure 5 The second decomposition signal diagram obtained by the SSEWT algorithm in the first experimental example of the present application;

[0047] Figure 6 The third decomposition signal diagram obtained by the SSEWT algorithm in the first experimental example of the present application;

[0048] Figure 7 The result comparison diagram of the SSEWT algorithm and the algorithm Rényi entropy in the first experimental example of the present application;

[0049] Figure 8 The instantaneous frequency diagram of the first decomposition signal after the Hilbert transform in the first experimental example of the present application;

[0050] Figure 9 The instantaneous frequency diagram of the first decomposition signal after the EE demodulation in the first experimental example of the present application;

[0051] Figure 10 The instantaneous frequency diagram of the second decomposition signal after the Hilbert transform in the first experimental example of the present application;

[0052] Figure 11 The instantaneous frequency diagram of the second decomposition signal after the EE demodulation in the first experimental example of the present application;

[0053] Figure 12 The oscillation curve diagram after the system is disturbed in the second experimental example of the present application;

[0054] Figure 13This is a schematic diagram of the decomposed signals IMF1 to IMF5 of the measured signal processed by the FSST algorithm in Experimental Example 2 of this invention;

[0055] Figure 14 This is a schematic diagram of the decomposed signals IMF1 to IMF3 of the measured signal processed by the SSEWT algorithm in Experimental Example 2 of this invention;

[0056] Figure 15 This is a schematic diagram of the system structure of the hybrid oscillation signal parameter identification system of the present invention.

[0057] Symbol explanation:

[0058] 1-Data acquisition module; 2-Signal decomposition module; 3-Signal judgment module; 4-Mode determination module. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] The purpose of this invention is to provide a method and system for identifying parameters of a hybrid oscillation signal, which can accurately analyze the characteristic parameters of the signal under test and thus accurately identify the dominant oscillation mode of the signal.

[0061] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] like Figure 1 As shown, an embodiment of the present invention provides a method for identifying parameters of a hybrid oscillation signal, comprising:

[0063] Step 100: Obtain the mixed oscillation signal.

[0064] Step 200: The mixed oscillation signal is decomposed using the second-order synchronous extraction wavelet transform method to obtain the decomposed signal.

[0065] Step 300: Determine whether the decomposed signal is completely decomposed based on the loss accuracy.

[0066] Step 400: Perform empirical envelope demodulation on the fully decomposed signal to obtain the dominant oscillation mode of the fully decomposed signal; the dominant oscillation mode includes low-frequency oscillation and ultra-low-frequency oscillation.

[0067] wherein in step 100, the mixed oscillation signal of the frequency of the hydroelectric generating set is generally non-stationary and nonlinear, and in the embodiment, the oscillation signal is assumed to be:

[0068]

[0069] wherein x k (t) represents the kth AM-FM oscillation component; A k (t) represents the instantaneous amplitude of the kth AM-FM component and A k (t) > 0; φ k (t) represents the phase function of the instantaneous frequency, and φ k (t) > 0; φ' k (t) = ω c is the instantaneous frequency; φ k0 is the initial phase angle; and δ(t) represents noise and other components.

[0070] As a preferred embodiment, step 200 specifically comprises:

[0071] First, the formula expression of the synchroextracting wavelet transform (SEWT) is obtained.

[0072] The wavelet transform of the signal x(t) is expanded on the time-scale plane by using the wavelet base function, and is represented as:

[0073]

[0074] wherein a is a scale factor for controlling the stretching amount of ψ(t), ψ(t) is a wavelet base function, is the complex conjugate of the wavelet base function, R is a real number set, x(u) is a signal expression, t is time, and u is a variable. Wherein ψ(t) is defined as:

[0075]

[0076] wherein σ is a parameter for controlling the shape of the wavelet base function.

[0077] The Fourier transform of x(t) and ψ(t) in equation (2) can be obtained as:

[0078]

[0079] wherein and respectively represent the Fourier transform of x(t) and ψ(t), Need to meet η is frequency, ω c is the center frequency.

[0080] In combination with formula (2) and formula (4), according to Plancherel theorem and the properties of Fourier transform, the frequency domain form of wavelet transform (WT) can be expressed as:

[0081]

[0082] In the formula, is the time-frequency coefficient under the frequency. Taking the partial derivative of both sides of formula (5) with respect to time t, the instantaneous frequency estimation value

[0083]

[0084] In the formula, Re[·] is the real part of the complex number, is the partial derivative.

[0085] Introducing Dirac function, only keeping the time-frequency coefficient under the real frequency, the synchroextracting wavelet transform (SEWT) can be defined as:

[0086]

[0087] In the formula, is the synchroextracting operator (SEO), satisfying:

[0088]

[0089] Secondly, the formula expression of the second-order synchroextracting wavelet transform (SSEWT) is obtained.

[0090] From formula (7), it can be seen that at this time WTe x (t,η) is determined by the instantaneous frequency . For If there is a small enough ε, satisfying |A′ k (t)|<ε, |φ″ k (t)|≤ε, according to the second-order Taylor expansion formula, the AM-FM component can be expressed as:

[0091]

[0092] In order to solve at this time First, the following formula needs to be calculated Substitute equation (9) and equation (3) into equation (2), we have:

[0093]

[0094] Where τ is the integral variable.

[0095] Where:

[0096]

[0097] Substitute equation (10) and equation (11) into equation (6), the instantaneous frequency can be expressed as:

[0098]

[0099] Further expressed as:

[0100]

[0101] Equation (13) shows that if the second derivative of the phase function is considered, The instantaneous frequency of the signal to be measured cannot be accurately represented. In order to further improve the instantaneous frequency estimation accuracy of SEWT, the second-order instantaneous frequency estimation quantity is used to obtain the time-frequency representation result, first calculate the wavelet transform of signal (9)

[0102]

[0103] In order to make the instantaneous frequency estimation value closer to the true frequency, this embodiment introduces the local time delay operator And the local modulation operator These two correction terms correct the instantaneous frequency, which is expressed as:

[0104]

[0105] Where The partial derivative with respect to time t is:

[0106]

[0107] Substitute equation (10), equation (14) and equation (16) into equation (15) to obtain:

[0108]

[0109] Equation (17) shows that the second derivative of the phase function at this time can be replaced by the local modulation operator Further, equation (13) is rewritten as:

[0110]

[0111] Thus the second-order instantaneous frequency estimate may be expressed as:

[0112]

[0113] may be expressed as:

[0114]

[0115] As a preferred embodiment, step 300 specifically comprises:

[0116] Firstly, reconstructing a plurality of the decomposition signals to obtain a reconstructed signal.

[0117] The reconstructed signal can be approximately reconstructed as:

[0118]

[0119] Secondly, judging whether the loss accuracy of the reconstructed signal is less than a preset threshold.

[0120] In order to accurately obtain all the decomposition signals (i.e. a plurality of IMF components) contained in the signal and prevent incomplete decomposition or over-decomposition, the loss accuracy (LA) is used as a measure of the similarity between the original signal and the reconstructed signal in the embodiment, and the calculation formula of the loss accuracy is:

[0121]

[0122] In the formula, x is the original input signal; ∑x k is the reconstructed signal; is a two-norm operator.

[0123] Thirdly, if yes (i.e. the loss accuracy of the reconstructed signal is less than the preset threshold), the decomposition signal is completely decomposed, and the completely decomposed decomposition signal is output. If no (i.e. the loss accuracy of the reconstructed signal is not less than the preset threshold), the decomposition signal is not completely decomposed, and the step of decomposing the mixed oscillation signal by using the second-order synchronous extraction wavelet transform method to obtain the decomposition signal is re-executed.

[0124] As a preferred embodiment, step 400 specifically comprises:

[0125] Firstly, performing empirical envelope demodulation on the completely decomposed decomposition signal to obtain mode information; the mode information at least includes amplitude, frequency, attenuation factor and damping ratio, and specifically includes:

[0126] S1: maximum operation is performed on the completely decomposed decomposition signal to obtain a maximum value and a corresponding time.

[0127] S2: a cubic spline function is used to fit the maximum value and the corresponding time to obtain an empirical envelope function.

[0128] S3: the completely decomposed decomposition signal is normalized according to the empirical envelope function to obtain a pure frequency modulation function.

[0129] S4: mode information is determined according to the completely decomposed decomposition signal and the pure frequency modulation function.

[0130] In the second step, a dominant mode of oscillation of the completely decomposed decomposition signal is determined according to the mode information.

[0131] In the first step, the mixed oscillation signal is decomposed by using the SSEWT to obtain K decomposition signals IMF components with limited bandwidths. In order to extract the instantaneous characteristics of the decomposition signal, an empirical envelope method (EE) is used to obtain comprehensive mode information of the oscillation signal, i.e. signal characteristic parameters such as amplitude, frequency, attenuation factor and damping ratio. As shown in the following formula: Figure 2

[0132] Decomposition signal x k ((t) can be represented as:

[0133]

[0134] In the formula, A k (t) is an instantaneous amplitude; φ k (t) is an instantaneous phase; A k0 is an initial amplitude of the kth oscillation signal; γ k is an attenuation factor; m k is an oscillation frequency; φ k0 is an initial phase angle.

[0135] The maximum value x h of |x(t)| and the corresponding time t h (h=1, 2, …, N) are obtained by replacing the envelope of x(t) with the envelope function a(t) of |x(t)| (where a(t)>0). h h (h=1, 2, …, N) are fitted using a cubic spline function to obtain an empirical envelope function a 11 (t) of the signal. 11 (t) is normalized until x 1n (t) is a pure frequency modulation function, i.e. a​​1n (t)≤1, iteration is stopped.

[0136]

[0137] recorded in such that:

[0138]

[0139] Taking derivative on both sides of the equation:

[0140]

[0141] where: Since the relative carrier changes slowly, so can be regarded as F′ k (t) of the envelope, formula (19) is repeated for F′ k (t), and we get:

[0142]

[0143] Further, the amplitude and frequency of the signal x k (t) can be obtained:

[0144]

[0145] The attenuation factor γ k can be obtained from formula (19) and formula (24), and the damping ratio of the corresponding mode is obtained by using the calculated γ k , m k , that is:

[0146]

[0147] The embodiment also provides two experimental examples for verification.

[0148] Experimental Example 1:

[0149] The self-synthesized signal is decomposed and compared. The oscillation frequency of the low-frequency oscillation signal is less than 2.5 Hz, the oscillation frequency of the ultra-low frequency oscillation signal is less than 0.1 Hz, and the frequency difference between different ultra-low frequency signal components is not obvious. In order to make the simulation signal more consistent with the actual hydroelectric generator system oscillation characteristics, the self-synthesized simulation test signal constructed in the embodiment contains two ultra-low frequency oscillation modes and one low-frequency oscillation mode.

[0150]

[0151] In the formula, δ(t) is the signal-to-noise ratio (SNR) of the original signal, the white Gaussian noise in the experiment is 10 dB, the sampling frequency is 100 Hz, the window parameter σ of the SSEWT is set to 5, and the waveform of the analog signal is as shown in Figure 3

[0152] Referring to Figures 4-6 Fig. 3, the SSEWT can effectively avoid modal aliasing and false modes, that is, a same IMF component contains frequency components with a large scale difference, and ensure that the IMF component has a high fitting degree with the original modal signal, thereby laying a foundation for subsequent identification of the characteristic parameters of the oscillation signal.

[0153] Referring to Figure 7 Fig. 4, for a mixed oscillation signal, an important indicator for measuring the effectiveness of a time-frequency analysis method is time-frequency concentration, that is, the higher the time-frequency concentration, the narrower the time-frequency curve. The SSEWT uses the Rényi entropy to quantitatively determine the degree of energy accumulation, and the Rényi entropy value is used to evaluate the time-frequency concentration. The smaller the Rényi entropy value, the higher the concentration degree of the function. Gaussian white noise with 0 dB, 10 dB, 20 dB, 30 dB, 40 dB, and 50 dB is added to the signal, and the Rényi entropy values of various algorithms under different signal-to-noise ratios are compared. α = 3 is used to quantitatively evaluate the time-frequency energy concentration degree of different methods. The Gaussian window function of the Fourier synchrosqueezed transform (FSST) and the short-time Fourier transform (STFT) is The window parameter σ is set to 0.02. It can be found from Figure 7 that the Rényi entropy always decreases with the decrease of the noise, which means that the noise weakens the energy concentration of the time-frequency analysis result. Under certain conditions, the effect of the SEWT is better than that of the STFT and the FSST. In the enlarged red box diagram, the intersection of the FSST and the SSEWT Rényi entropy curves can be seen. On the right side of the intersection, the SSEWT Rényi entropy curve is always below the FSST, indicating that the time-frequency expression of the SSEWT is better than that of the FSST. Although the low-order synchrosqueezing transform is relatively simple for a system, when the noise is high, the SEWT cannot meet the accuracy requirement, and the SSEWT needs to be further applied to accurately decompose the signal, which shows that the method has strong robustness.

[0154] Referring to Figures 8-11 ​The results of the instantaneous frequency obtained by Hilbert transform and EE demodulation are shown in FIG. 2. It can be seen from the figure that both methods can obtain the instantaneous frequency of the signal, but EE demodulation can avoid the energy leakage at the endpoints caused by Hilbert transform, so EE demodulation can well suppress the endpoint effect and has more accurate estimation performance.

[0155] Table 1 is a comparison of the mode recognition results of each modal signal and the theoretical value. The recognition results show that the dominant mode of oscillation can be accurately separated from the complex signal by SSEWT, and EE demodulation technology has high recognition degree for the decomposed IMF components. The amplitude, frequency, attenuation factor and damping ratio of each IMF component are more accurate than the theoretical value.

[0156] Table 1 Comparison of modal characteristic parameters and theoretical values

[0157]

[0158]

[0159] Example two:

[0160] After asynchronous networking, the southwest power grid will operate as an independent sending end network with a total installed capacity of 120 million kW, of which the total installed capacity of hydropower is about 85 million kW, accounting for about 70% of the total installed capacity. It can be seen that the proportion of hydropower units in the southwest power grid is very high, which is more prone to frequency oscillation. Taking the PMU measurement data of a DC island system in the southwest power grid as an example, the feasibility and effectiveness of SSEWT in identifying the modal parameters of the oscillation signal are verified. When the system is disturbed, the oscillation curve is shown in FIG. 2. The oscillation signal contains three groups of oscillation modes: x1(t), x2(t) and x3(t), and the frequency and damping ratio are 1.43 Hz, -0.0222; 0.064 Hz, 0.0249 and 0.018 Hz, 0.0618, respectively. Figure 12 For reference

[0161] and Figure 13 and Figure 14It can be seen that the FSST decomposes five dominant components, the modal components 1-2 are not well removed from noise in the decomposition process, and the noise is mixed in the decomposition result; the signal frequency in the modal component 3 is not unique, and there is a certain degree of modal aliasing phenomenon; the modal components 4-5 are decomposed into two or more components due to the inability of the FSST to accurately identify the frequency components of similar characteristic time scales. The signal components decomposed by the two algorithms are quite different in quantity and form. From the above analysis, it is further proved that the method of the embodiment has good processing accuracy for time-varying and frequency-close nonlinear signals, and can accurately obtain the characteristic information of each oscillation component in the mixed signal.

[0162] The specific values are identified as shown in Table 2. It can be seen that the method of the application has a small error and has strong parameter identification ability for complex and frequency-close oscillation signals, while the processing ability of the STFT algorithm for complex signals is obviously insufficient.

[0163] Table 2 Comparison of identification results

[0164]

[0165]

[0166] The application proposes to apply a second-order synchronous extraction transform to pre-process the collected complex oscillation signal. The one-dimensional time domain signal is expanded to a two-dimensional time-frequency surface, only the time-frequency coefficients of the signal position in the two-dimensional time-frequency surface are retained, the noise and other components irrelevant to the true frequency are eliminated, the remaining coefficients are recombined to obtain a new two-dimensional time-frequency curve. The interference of noise and the modal aliasing phenomenon are avoided, the accuracy in time and frequency is met, the characteristic information of the signal to be analyzed can be more clearly observed and extracted, and good signal reconstruction ability is maintained. And the EE demodulation is used to identify the parameters of a series of single-component signals decomposed, the complex integral transform process is saved, the end effect is avoided, the amplitude, frequency, attenuation factor and damping ratio and other characteristic parameters of the signal to be measured can be more accurately obtained, and accurate identification of the signal is realized.

[0167] Referring to Figure 15 The application further discloses a mixed oscillation signal parameter identification system, which comprises:

[0168] A data acquisition module 1 is used to acquire a mixed oscillation signal;

[0169] A signal decomposition module 2 is used to decompose the mixed oscillation signal by using a second-order synchronous extraction wavelet transform method to obtain a decomposed signal;

[0170] The signal judging module 3 is configured to determine whether the decomposition signals are completely decomposed according to the loss precision.

[0171] The modal determining module 4 is configured to perform empirical envelope demodulation on the completely decomposed decomposition signals to obtain oscillation dominant modes of the completely decomposed decomposition signals; the oscillation dominant modes include low-frequency oscillation and ultra-low-frequency oscillation.

[0172] In a further aspect, the signal judging module includes:

[0173] The signal reconstructing unit is configured to reconstruct the plurality of decomposition signals to obtain a reconstructed signal.

[0174] The signal judging unit is configured to determine whether the loss precision of the reconstructed signal is less than a preset threshold.

[0175] The signal output unit is configured to output the completely decomposed decomposition signals when the loss precision of the reconstructed signal is less than the preset threshold.

[0176] In a further aspect, the modal determining module includes:

[0177] The mode information determining unit is configured to perform empirical envelope demodulation on the completely decomposed decomposition signals to obtain mode information; the mode information at least includes amplitude, frequency, attenuation factor and damping ratio.

[0178] The modal determining unit is configured to determine the oscillation dominant modes of the completely decomposed decomposition signals according to the mode information.

[0179] In a further aspect, the mode information determining unit includes:

[0180] The data operation sub-unit is configured to perform maximum value operation on the completely decomposed decomposition signals to obtain maximum values and corresponding time points.

[0181] The data fitting sub-unit is configured to fit the maximum values and the corresponding time points by using a cubic spline function to obtain an empirical envelope function.

[0182] The standardization processing sub-unit is configured to perform standardization processing on the completely decomposed decomposition signals according to the empirical envelope function to obtain a pure frequency modulation function.

[0183] The mode information determining sub-unit is configured to determine mode information according to the completely decomposed decomposition signals and the pure frequency modulation function.

[0184] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0185] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for identifying parameters of a hybrid oscillation signal, characterized in that, include: Acquire mixed oscillation signal; The hybrid oscillation signal is decomposed using a second-order synchronous extraction wavelet transform method to obtain the decomposed signal; Whether the decomposed signal is completely decomposed is determined based on the loss accuracy. Empirical envelope demodulation is performed on the fully decomposed signal to obtain the dominant oscillation mode of the fully decomposed signal. The dominant oscillation modes include low-frequency oscillations and ultra-low-frequency oscillations; The second-order synchronous extraction wavelet transform method includes: Obtain the formulaic expression for synchronous extraction of wavelet transform, specifically including: Using wavelet basis functions, the signal is expanded by wavelet transform in the time-scale plane, as follows: Where 'a' is the scaling factor, used to control... The amount of expansion and contraction, For wavelet basis functions, Let R be the complex conjugate of the wavelet basis functions, and R be the set of real numbers. For signal expression, t For time, u As the independent variable; Perform Fourier transform on the signal and the wavelet basis functions; Based on Passevar's theorem and the properties of the Fourier transform, the frequency domain form of the wavelet transform is obtained, expressed as: in, and They represent and Fourier transform, Must meet , For frequency These are the time-frequency coefficients at that frequency; By taking the partial derivatives with respect to time on both sides of the frequency domain form, the instantaneous frequency estimate is obtained. : in, , Let be the real part of the complex number. These are partial derivatives; Introducing the Dirac function, which retains only the time-frequency coefficients corresponding to the true frequency, the synchronous extraction wavelet transform is defined as follows: in, To extract operators synchronously, the following conditions must be met: ； The formula for obtaining the second-order synchronous extraction wavelet transform is as follows: The AM-FM components are represented by a second-order Taylor expansion: ； Calculate the wavelet transform of the signal; introduce a local time delay operator. and local modulation operator The correction for the instantaneous frequency is expressed as: ； Based on the corrected instantaneous frequency, the second-order instantaneous frequency estimate is expressed as follows: ； Based on the second-order instantaneous frequency estimate, the formula for the second-order synchronous extraction wavelet transform is obtained, expressed as: 。 2. The method for identifying parameters of a hybrid oscillation signal according to claim 1, characterized in that, The step of determining whether the decomposed signal is completely decomposed based on the loss accuracy specifically includes: The decomposed signals are reconstructed to obtain the reconstructed signals; Determine whether the loss accuracy of the reconstructed signal is less than a preset threshold; If so, the decomposed signal is completely decomposed, and the fully decomposed signal is output.

3. The method for identifying parameters of a hybrid oscillation signal according to claim 1, characterized in that, The step of performing empirical envelope demodulation on the fully decomposed signal to obtain the dominant oscillation mode of the fully decomposed signal specifically includes: The fully decomposed signal is subjected to empirical envelope demodulation to obtain mode information; the mode information includes at least amplitude, frequency, attenuation factor and damping ratio. The dominant oscillation mode of the fully decomposed signal is determined based on the mode information.

4. The method for identifying parameters of a hybrid oscillation signal according to claim 3, characterized in that, The empirical envelope demodulation of the fully decomposed signal to obtain mode information specifically includes: Perform a maximum value operation on the fully decomposed signal to obtain the maximum value and the corresponding time. An empirical envelope function is obtained by fitting the maximum value and the corresponding time point using a cubic spline function; The fully decomposed signal is standardized according to the empirical envelope function to obtain a pure frequency modulation function; The mode information is determined based on the fully decomposed signal and the pure frequency modulation function.

5. A hybrid oscillation signal parameter identification system, characterized in that, include: The data acquisition module is used to acquire the mixed oscillation signal; The signal decomposition module is used to decompose the hybrid oscillation signal using a second-order synchronous extraction wavelet transform method to obtain the decomposed signal; The signal determination module is used to determine whether the decomposed signal has been completely decomposed based on the loss accuracy. The mode determination module is used to perform empirical envelope demodulation on the fully decomposed signal to obtain the dominant oscillation mode of the fully decomposed signal. The dominant oscillation modes include low-frequency oscillations and ultra-low-frequency oscillations; The second-order synchronous extraction wavelet transform method includes: Obtain the formulaic expression for synchronous extraction of wavelet transform, specifically including: Using wavelet basis functions, the signal is expanded by wavelet transform in the time-scale plane, as follows: Where 'a' is the scaling factor, used to control... The amount of expansion and contraction, For wavelet basis functions, Let R be the complex conjugate of the wavelet basis functions, and R be the set of real numbers. For signal expression, t For time, u As the independent variable; Perform Fourier transform on the signal and the wavelet basis functions; Based on Passevar's theorem and the properties of the Fourier transform, the frequency domain form of the wavelet transform is obtained, expressed as: in, and They represent and Fourier transform, Must meet , For frequency These are the time-frequency coefficients at that frequency; By taking the partial derivatives with respect to time on both sides of the frequency domain form, the instantaneous frequency estimate is obtained. : in, , Let be the real part of the complex number. These are partial derivatives; Introducing the Dirac function, which retains only the time-frequency coefficients corresponding to the true frequency, the synchronous extraction wavelet transform is defined as follows: in, To extract operators synchronously, the following conditions must be met: ； The formula for obtaining the second-order synchronous extraction wavelet transform is as follows: The AM-FM components are represented by a second-order Taylor expansion: ； Calculate the wavelet transform of the signal; introduce a local time delay operator. and local modulation operator The correction for the instantaneous frequency is expressed as: ； Based on the corrected instantaneous frequency, the second-order instantaneous frequency estimate is expressed as follows: ； Based on the second-order instantaneous frequency estimate, the formula for the second-order synchronous extraction wavelet transform is obtained, expressed as: 。 6. The hybrid oscillation signal parameter identification system according to claim 5, characterized in that, The signal determination module includes: A signal reconstruction unit is used to reconstruct several of the decomposed signals to obtain a reconstructed signal; A signal determination unit is used to determine whether the loss accuracy of the reconstructed signal is less than a preset threshold. The signal output unit is used to output a fully decomposed signal when the loss accuracy of the reconstructed signal is less than a preset threshold.

7. The hybrid oscillation signal parameter identification system according to claim 5, characterized in that, The modality determination module includes: The mode information determination unit is used to perform empirical envelope demodulation on the fully decomposed signal to obtain mode information; the mode information includes at least amplitude, frequency, attenuation factor and damping ratio. A mode determination unit is used to determine the dominant oscillation mode of the fully decomposed signal based on the mode information.

8. The hybrid oscillation signal parameter identification system according to claim 7, characterized in that, The pattern information determination unit includes: The data processing subunit is used to perform maximum value operation on the fully decomposed signal to obtain the maximum value and the corresponding time. The data fitting subunit is used to fit the maximum value and the corresponding time using a cubic spline function to obtain an empirical envelope function; The standardization processing subunit is used to standardize the fully decomposed signal according to the empirical envelope function to obtain a pure frequency modulation function. The mode information determination subunit is used to determine mode information based on the fully decomposed signal and the pure frequency modulation function.

Images