Rapid high-precision frequency measurement method for super-synchronous inter-harmonics

Abstract

The invention discloses a rapid high-precision frequency measurement method for super-synchronous inter-harmonics, and relates to the field of signal processing, and the method comprises the steps: S1, presetting a reference fundamental wave signal, and generating a modulation fundamental wave signal on a discrete time sequence through a super-synchronous source; s2, setting the reference fundamental wave signal and the modulation fundamental wave signal in a discrete time sequence, and intercepting an abnormal frequency interval; s3, extracting all abnormal fundamental wave data and all reference fundamental wave data; s4, abnormal components, higher than the corresponding reference fundamental wave data, of the abnormal fundamental wave data are screened out, and accurate frequency harmonic correction is carried out. The blind source separation algorithm is adopted to accurately extract the abnormal frequency interval, and the abnormal fundamental wave data and the reference fundamental wave data are compared, so that the frequency correction can be carried out in time, the data processing and analysis can be carried out in real time, the processing delay is reduced, and the reliability of the frequency measurement is improved.

Description

Technical Field

[0001] The invention relates to the field of signal processing, and in particular to a fast and high-precision frequency measurement method for super-synchronous interharmonics. Background Art

[0002] With the widespread application of power electronic equipment, the generation of supersynchronous interharmonics has become an important issue. These harmonic components with frequencies higher than the fundamental wave will lead to the appearance of oscillation components, affecting the dynamic stability and safety of the system. Supersynchronous interharmonics refer to harmonic components with frequencies higher than the fundamental frequency generated in the interaction between power electronic equipment and the power grid. These harmonics are usually caused by power electronic devices (such as inverters, rectifiers, etc.) under switching operations or nonlinear loads. Since the frequency of supersynchronous interharmonics is close to the fundamental frequency and its low-order harmonics (such as second and third harmonics), it is difficult to effectively separate these frequency components in spectrum analysis, resulting in a frequency overlap effect. If the overlapping high-frequency harmonic components have a large amplitude, the measured fundamental amplitude may be high, resulting in an abnormal increase in the corresponding parameters (such as power, effective value). The existing high-precision frequency separation algorithms usually lack real-time processing and are difficult to apply to rapidly changing power system environments. When processing abnormal harmonics of the signal, the accuracy of capturing the abnormal part of the signal is insufficient, resulting in difficulty in local filtering of the signal. Therefore, in view of the frequency overlap of oscillation components, a frequency measurement method with relatively lower cost and more immediate and efficient processing is needed in the prior art. Summary of the invention

[0003] The present invention provides a fast and high-precision frequency measurement method for supersynchronous interharmonics, which solves the problems of lack of accuracy and poor real-time local processing capability when measuring the frequency of intermittent wave oscillation components in the prior art.

[0004] The present invention is achieved through the following technical solutions: A fast and high-precision frequency measurement method for super-synchronous interharmonics, the method comprising: Step S1: preset a reference fundamental wave signal, use a super-synchronous source to generate an initial fundamental wave signal on a discrete time series, use a nonlinear element to harmonically modulate the initial fundamental wave signal and output a modulated fundamental wave signal indicating that super-synchronization is completed; Step S2: setting the reference fundamental wave signal and the modulated fundamental wave signal in discrete time series, screening out the discrete signal components with abnormal frequencies in the modulated fundamental wave signal and marking them as abnormal components, and using a blind source separation algorithm to extract the abnormal frequency interval with frequency overlap in each abnormal component; Step S3: using an envelope detection method to extract abnormal fundamental wave data existing in all abnormal frequency intervals, and extracting all reference fundamental wave data in the reference fundamental wave signal; Step S4: Compare all abnormal fundamental wave data with the corresponding reference fundamental wave data at the same time series point, filter out the abnormal components of the abnormal fundamental wave data that are higher than the corresponding reference fundamental wave data, and perform precise frequency harmonic correction.

[0005] Since the frequency of super-synchronous interharmonics is close to the fundamental frequency and its low-order harmonics (such as second and third harmonics), it is difficult to effectively separate these frequency components in spectrum analysis, resulting in a frequency overlap effect. If the overlapping high-frequency harmonic components have a large amplitude, the measured fundamental amplitude may be too high, causing the corresponding parameters (such as power and effective value) to increase abnormally. The currently existing high-precision frequency separation algorithms usually lack real-time processing and are difficult to apply to rapidly changing power system environments. When processing abnormal harmonics of the signal, the accuracy of capturing the abnormal part of the signal is insufficient, resulting in greater difficulty in local filtering of the signal. Based on this, the present invention provides a fast and high-precision frequency measurement method for super-synchronous interharmonics to solve the problems of lack of accuracy and poor real-time local processing capabilities when measuring the frequency of the intermittent wave oscillation component in the prior art.

[0006] Furthermore, the blind source separation algorithm intercepts the abnormal frequency interval content including: setting a power threshold for screening out abnormal frequency values, using the FastICA analysis method to maximize the non-Gaussian method to extract the independent fundamental signal representing the existence of overlapping frequencies in the abnormal component, performing FFT transformation on the independent fundamental signal and marking the output spectrum as a modulation spectrum, marking the frequency value corresponding to the power peak value in the modulation spectrum that is higher than the power threshold as an abnormal frequency value; and calculating the abnormal frequency interval representing the frequency overlapping area based on the half-peak width of the abnormal frequency value.

[0007] Furthermore, the abnormal frequency interval is calculated using a sliding window method based on the half-peak width of the abnormal frequency value, and the process includes: setting the sliding window size based on the frequency resolution of the independent fundamental signal, initializing the sliding window from each abnormal frequency value to both ends of the frequency value axis, extracting the frequency components and corresponding power values ​​within the current window range at each sliding window position, and identifying the power peak within the window; When the power peak in the sliding window on both sides of the abnormal frequency value exceeds the power threshold, the sliding window slides one unit toward both ends of the frequency value axis; when the power peak in the sliding window on both sides of the abnormal frequency value is less than or equal to the power threshold, the frequency value corresponding to the power peak in the sliding window on both sides of the abnormal frequency value is set as the upper and lower boundary values ​​of the abnormal frequency interval.

[0008] Furthermore, the process of setting the sliding window size includes: assuming that the frequency resolution of the independent fundamental wave signal required by the sliding window is represented by Δf, assuming that the sampling frequency of the independent fundamental wave signal is represented by f s, let the number of FFT transformation points be N, the sliding window size be W, and set the adjustment constant k, Then the calculation formula of the sliding window W is expressed as: .

[0009] Furthermore, a white noise value for adding to FastICA analysis is set based on the modulation spectrum, and the calculation and setting process of the white noise value includes: calculating the mean of the frequency values ​​in the modulation spectrum, and calculating the covariance matrix after subtracting the mean from each data point, performing eigenvalue decomposition on the covariance matrix and performing an inverse square root transformation to output the white noise value.

[0010] Furthermore, the specific contents of the envelope detection method include: using Hilbert transform to extract the signal envelope value of the abnormal component; performing peak detection on the envelope signal to identify the significant peak in the envelope signal, after screening the corresponding time point on the envelope signal, extracting the corresponding abnormal fundamental wave data from the abnormal component according to the detected peak position.

[0011] Furthermore, the process of extracting the signal envelope value of the abnormal component using Hilbert transform includes: acquiring the instantaneous signal of the abnormal component using Hilbert transform, and calculating the signal envelope value based on the instantaneous signal.

[0012] Furthermore, the current time point of the design calculation is represented as t, and the integral variable is represented as Used to represent different time points in the original signal; suppose the original signal of the abnormal component is represented as , the instantaneous signal is represented by x h (t), Then the instantaneous signal x h The calculation formula of (t) is expressed as: .

[0013] Furthermore, let the signal envelope value be represented as E(t), and let the original signal at time t be represented as x(t), Then the calculation formula of the signal envelope value E(t) is expressed as: .

[0014] Furthermore, the Hilbert transformed results and the extracted fundamental data are visualized, and a smoothing filter is applied on the calculated signal envelope values.

[0015] Compared with the prior art, the present invention has the following advantages and beneficial effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings: Figure 1 It is a schematic diagram of the structure of the present invention. DETAILED DESCRIPTION

[0017] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention. Example

[0018] like Figure 1 As shown, this embodiment is a fast and high-precision frequency measurement method for super-synchronous interharmonics, the method comprising: Step S1: preset a reference fundamental wave signal, use a super-synchronous source to generate an initial fundamental wave signal on a discrete time series, use a nonlinear element to harmonically modulate the initial fundamental wave signal and output a modulated fundamental wave signal indicating that super-synchronization is completed; Step S2: setting the reference fundamental wave signal and the modulated fundamental wave signal in discrete time series, screening out the discrete signal components with abnormal frequencies in the modulated fundamental wave signal and marking them as abnormal components, and using a blind source separation algorithm to extract the abnormal frequency interval with frequency overlap in each abnormal component; Step S3: using an envelope detection method to extract abnormal fundamental wave data existing in all abnormal frequency intervals, and extracting all reference fundamental wave data in the reference fundamental wave signal; Step S4: Compare all abnormal fundamental wave data with the corresponding reference fundamental wave data at the same time series point, filter out the abnormal components of the abnormal fundamental wave data that are higher than the corresponding reference fundamental wave data, and perform precise frequency harmonic correction.

[0019] The fundamental signal refers to the sine wave component with the lowest frequency in the periodic signal. It is the main component of the power electronic device signal, and other frequency components, such as harmonics or noise, are integer multiples of the fundamental signal. The reference fundamental signal refers to a stable signal used as a reference in signal processing and measurement, usually a signal of known frequency and amplitude, used to detect and measure the frequency of the target signal, and to judge the frequency deviation of the signal by comparing it with the reference fundamental signal. The super-synchronous source refers to a signal source that can generate a very stable and accurate frequency, such as used to generate the initial fundamental signal in this embodiment, and is usually used to synchronize multiple signals; when the fundamental signal passes through a nonlinear element, in addition to retaining the original frequency, it also generates harmonic signals with integer multiple frequencies. These harmonic signals form a modulated signal together with the fundamental signal. The modulated fundamental signal refers to the fundamental signal generated by the super-synchronous source, which can be synchronized with the reference fundamental signal within a specific frequency range after harmonic modulation. Super-synchronization means that the modulated signal is highly consistent with the reference signal in frequency and phase, so that the signal forms an ideal waveform in the frequency domain, which is convenient for subsequent frequency measurement and analysis. The discrete time series form represents the conversion of a continuous signal into a series of discrete numerical samples. In specific applications, the frequency anomaly in the modulated fundamental signal can be judged based on the characteristics of frequency offset anomaly, frequency amplitude anomaly, harmonic frequency multiplication relationship anomaly, frequency phase anomaly, etc. The purpose of the blind source separation algorithm is to separate the original signal from multiple mixed signals; the abnormal frequency interval refers to the area where multiple frequency components overlap in the abnormal signal. These overlapping parts may lead to inaccurate measurement results. Through blind source separation, each independent abnormal frequency component can be extracted from the mixed signal, providing clear information for subsequent frequency analysis.

[0020] More importantly, the fundamental wave data mainly includes conventional signal data contents such as frequency components, amplitude phase, and time characteristics; the abnormal fundamental wave data refers to signal components detected within a specific frequency range that are significantly different from the reference fundamental wave signal. These data usually indicate potential problems or abnormal phenomena in the system. The envelope detection method is a signal processing technology that identifies the amplitude change of the signal by extracting the envelope of the signal; it is usually used to process modulated signals or non-stationary signals, and can effectively extract the main features of the signal. All abnormal fundamental wave data are compared with the corresponding reference fundamental wave data, and the comparison content is mainly based on the consistency comparison of amplitude changes, frequency analysis, phase offset and noise evaluation in conventional technologies; comparison at the same time series point is to ensure that the abnormal fundamental wave data and the reference fundamental wave data can maintain consistency in the time dimension. The precise frequency harmonic correction can be completed by phase-locked loop, DSP processing or harmonic precise compensation based on existing technologies to correct abnormal harmonics.

[0021] Furthermore, as a feasible implementation method, the blind source separation algorithm intercepts the abnormal frequency interval content including: setting a power threshold for screening out abnormal frequency values, using the FastICA analysis method to maximize the non-Gaussianity to extract the independent fundamental signal representing the existence of overlapping frequencies in the abnormal component, performing FFT transformation on the independent fundamental signal and marking the output spectrum as a modulation spectrum, marking the frequency value corresponding to the power peak value in the modulation spectrum that is higher than the power threshold as an abnormal frequency value; and calculating the abnormal frequency interval representing the frequency overlapping area based on the half-peak width of the abnormal frequency value.

[0022] The power threshold is a preset value, which can be set based on the power spectrum density or energy level of the signal in combination with the high harmonic signal characteristics in this field. For example, the mean and standard deviation of the modulated spectrum can be calculated according to the power distribution of the modulated spectrum, and the threshold can be set to the mean plus a certain multiple of the standard deviation to ensure that only significant abnormal frequencies are screened out; or based on the threshold of spectrum analysis, the modulated spectrum is analyzed, and after the main components of the spectrum are identified, the threshold is set to a certain proportion of these main components, such as 70%, etc., to ensure that enough abnormal frequencies are identified. The initial fundamental signal is harmonically modulated using a nonlinear element. Since the output of the nonlinear element and the input signal are not a simple linear relationship, the input fundamental signal can generate additional frequency harmonic components after passing through the element. The FastICA analysis method is a fixed point-independent component analysis method, which is a kind of blind source separation algorithm. It is widely used in the field of signal processing for its fast convergence speed and good separation effect; the algorithm can well estimate the original signals that are statistically independent and mixed by unknown factors from the observed signal. Non-Gaussianity means that the probability distribution of a signal does not follow the Gaussian distribution and usually has high kurtosis and skewness. FastICA analysis can use this characteristic to distinguish different signal sources. By maximizing non-Gaussianity, it is possible to extract signal components that are statistically independent of each other. At the same time, non-Gaussianity can make the separation of overlapping frequency components more obvious, improve the separability of the signal, and thus enhance the accuracy of subsequent analysis. The independent fundamental signal refers to the fundamental signal that is independent of each other and not affected by other frequency components extracted from a signal containing overlapping frequencies during the blind source separation process. These signals are non-Gaussian in statistical characteristics and represent different frequency components, which can reflect the true characteristics of the source signal. The modulation spectrum refers to the spectrum data processed by fast Fourier transform (i.e., FFT transform), which reflects the distribution of each frequency component in the harmonically modulated signal. The half-peak width refers to the width between the corresponding two frequency points in the spectrum when the peak intensity is half. It reflects the distribution and intensity of the frequency components and can provide information about the shape of the spectrum. Combining the blind source separation algorithm with the characteristics of nonlinear elements, it can not only effectively extract and analyze the abnormal frequencies in supersynchronous interharmonics, but also provide strong support for high-precision frequency measurement through accurate calculation of frequency overlapping areas.

[0023] Further, as a feasible implementation method, the abnormal frequency interval is calculated using a sliding window method based on the half-peak width of the abnormal frequency value, and the process includes: setting the sliding window size based on the frequency resolution of the independent fundamental signal, initializing the sliding window from each abnormal frequency value to both ends of the frequency value axis, extracting the frequency components and corresponding power values ​​within the current window range at each sliding window position, and identifying the power peak value within the window; When the power peak in the sliding window on both sides of the abnormal frequency value exceeds the power threshold, the sliding window slides one unit toward both ends of the frequency value axis; when the power peak in the sliding window on both sides of the abnormal frequency value is less than or equal to the power threshold, the frequency value corresponding to the power peak in the sliding window on both sides of the abnormal frequency value is set as the upper and lower boundary values ​​of the abnormal frequency interval.

[0024] The frequency resolution refers to the minimum frequency interval that a signal can distinguish in the frequency domain. It is determined by the sampling frequency and the signal length, and is usually calculated using the formula Δf=f s / N means, where f s Represents the sampling frequency, and N represents the number of signal sampling points. The sliding window is initialized from each abnormal frequency value, that is, each identified abnormal frequency value is regarded as the center point of the sliding window. The sliding window will start from the abnormal frequency value and expand to both sides of the frequency value axis. The window will extend a certain range in the low frequency (left side) and high frequency (right side) directions to form an interval containing multiple frequency components. The extended range is determined by the previously set window size, which is usually determined by the frequency resolution of the signal. The size of the sliding window is usually matched with the frequency resolution. The larger the window, the more frequency components can be captured, but the accuracy may be reduced; the smaller the window, the more detailed frequency information can be provided, but some important components may be missed. The appropriate window size is selected according to the characteristics of the independent fundamental signal to balance the frequency resolution and signal analysis. For example, for an independent fundamental signal with a fast frequency change, a smaller window may be required to capture details; while for a slowly changing signal, a larger window can be selected. At each position of the sliding window, a frequency range consisting of multiple frequency components extracted by FFT analysis is defined, and each frequency component has a corresponding power value, indicating the intensity of the frequency component in the signal. The power value is usually the square of the amplitude of each frequency point in the spectrum, reflecting the energy distribution of the signal at this frequency. Extracting the power value helps to identify the main features of the signal. Within the current window range, find the point with the highest power value, and the frequency corresponding to this point is the power peak. The sliding window will slide to both sides of the frequency value axis to expand the abnormal frequency interval to ensure that no important frequency components are missed. The upper and lower boundary values ​​of the abnormal frequency interval are the frequency values ​​corresponding to the power peak in the left sliding window and the right sliding window of the abnormal frequency value, respectively.

[0025] Further, as a feasible implementation, the process of setting the sliding window size includes: assuming that the frequency resolution of the independent fundamental wave signal required by the sliding window is represented by Δf, assuming that the sampling frequency of the independent fundamental wave signal is represented by f s , let the number of FFT transformation points be N, the sliding window size be W, and set the adjustment constant k, Then the calculation formula of the sliding window W is expressed as: .

[0026] The frequency resolution Δf represents the interval between each frequency point in the frequency domain, which is usually determined by the sampling frequency of the signal and the number of FFT transformation points. The smaller the frequency resolution, the richer the details of the spectrum. s It indicates the frequency when the signal is sampled, which affects the time resolution of the signal and the frequency range that can be analyzed. The number of FFT transformation points N indicates the number of data points used when performing fast Fourier transform. The larger N is, the higher the spectrum resolution is, but the amount of calculation will also increase. The sliding window size W indicates the width of the sliding window, which mainly affects the range of frequency components that can be captured during analysis. The adjustment constant k is used to adjust the coefficient of the sliding window size, which can be set according to specific application requirements to optimize the analysis effect. The frequency resolution is determined by the sampling frequency and the number of FFT points; in the FFT transformation, the interval of each frequency point is proportional to the sampling frequency and inversely proportional to the number of FFT transformation points. Multiply the frequency resolution Δf by the adjustment constant k to obtain the size W of the sliding window. This size can be adjusted according to the specific signal characteristics and analysis requirements to ensure that important frequency components are captured. The selection of the adjustment constant k can affect the flexibility and adaptability of the sliding window. For example, choosing a larger k can make the window wider, which helps to capture a wider frequency range, while choosing a smaller k can improve the accuracy of frequency analysis.

[0027] Furthermore, as a feasible implementation method, a white noise value for adding to FastICA analysis is set based on the modulation spectrum, and the calculation and setting process of the white noise value includes: calculating the mean of the frequency values ​​in the modulation spectrum, and calculating the covariance matrix after subtracting the mean from each data point, performing eigenvalue decomposition on the covariance matrix and performing an inverse square root transformation to output the white noise value.

[0028] The white noise value is usually added in the preprocessing stage of FastICA analysis. In this stage, white noise is added to the input signal to enhance the independence of the signal and improve the performance of the algorithm. The white noise value refers to a frequency component with uniform distribution, whose power spectrum is flat over the entire frequency range, and is usually used as background noise or interference signal in signal processing. In FastICA analysis, adding white noise can help improve the independence and separation of signals. The mean reflects the overall intensity level of the modulation spectrum and can be used as a benchmark for data centering. The mean is subtracted from each data point, a process called centering. Through centering, the mean of the data becomes zero, which helps to eliminate the offset of the data and make the subsequent covariance calculation more accurate. The covariance matrix is ​​used to describe the relationship and change between the variables in the data. The covariance matrix is ​​subjected to eigenvalue decomposition to obtain eigenvalues ​​and eigenvectors. The eigenvalues ​​represent the degree of data variation in each direction, and the eigenvectors represent the main component directions of the data; the eigenvalues ​​are subjected to inverse square root transformation, which can adjust the scale of the eigenvalues ​​to a standardized range. By calculating the mean of the modulation spectrum and performing centering, covariance matrix construction and eigenvalue decomposition, combined with the inverse square root transformation process, a white noise value suitable for FastICA analysis can be effectively generated. This implementation improves the accuracy of signal processing and helps to extract more independent and clear components from complex signals.

[0029] Furthermore, as a feasible implementation method, the specific content of the envelope detection method includes: using Hilbert transform to extract the signal envelope value of the abnormal component; performing peak detection on the envelope signal to identify the significant peak in the envelope signal, after screening the corresponding time point on the envelope signal, extracting the corresponding abnormal fundamental wave data from the abnormal component according to the detected peak position; the process of extracting the signal envelope value of the abnormal component using Hilbert transform includes: using Hilbert transform to obtain the instantaneous signal of the abnormal component, and calculating the signal envelope value based on the instantaneous signal.

[0030] The Hilbert transform is a mathematical tool for generating functions with the same domain as functions. It is used for signal processing and can generate analytical signals from real signals. The analytical signal contains the amplitude and phase information of the original signal and can effectively extract the instantaneous characteristics of the signal. The abnormal component is processed using the Hilbert transform to generate its corresponding analytical signal. The real part of the analytical signal is the original signal, and the imaginary part is the result after the Hilbert transform. Peak detection is performed on the extracted envelope signal to identify significant peaks in the envelope signal; commonly used peak detection methods include threshold method and maximum value detection, the purpose of which is to find peaks in the envelope signal that can represent important features. According to the position of the detected envelope signal peak, the corresponding time points are screened out, which represent the moments when important features in the abnormal signal appear.

[0031] As a preferred implementation, the Hilbert transform result and the extracted fundamental wave data are visualized, and a smoothing filter is applied to the calculated signal envelope value. h (t) can be displayed graphically, usually with time as the horizontal axis and amplitude as the vertical axis, to help observe the changing characteristics of the instantaneous signal. The extracted fundamental wave data should also be visualized, using similar time-amplitude graphs to facilitate comparison with the instantaneous signal. This allows intuitive analysis of the main components of the signal and its changing trends. The main purpose of applying a smoothing filter to the signal envelope is to remove high-frequency noise, reduce random fluctuations in the signal, and make the envelope signal smoother and easier to analyze. By visualizing the Hilbert transform results and the extracted fundamental wave data, the signal characteristics can be intuitively analyzed, and smoothing the signal envelope values ​​helps improve the accuracy and readability of signal analysis.

[0032] The calculation process of the trend signal is defined as follows: the current time point of the calculation is represented as t, and the integral variable is represented as Used to represent different time points in the original signal; suppose the original signal of the abnormal component is represented as , the instantaneous signal is represented by x h (t), then the instantaneous signal x h The calculation formula of (t) is expressed as: . represents the weight factor, as When approaching t, The value becomes smaller, resulting in the original signal For the instantaneous signal x h (t) increases. This setting emphasizes the local characteristics of the instantaneous signal and highlights the relevance to the current moment. The entire calculation formula is calculated by The weighted integral of can effectively extract the instantaneous signal x at a specific time t h (t). This method helps to capture the dynamic characteristics of the signal and provide important data for subsequent analysis.

[0033] Furthermore, the calculation process of the signal envelope value is defined as follows: assuming that the signal envelope value is represented by E(t), and assuming that the original signal at time t is represented by x(t), then the calculation formula of the signal envelope value E(t) is expressed as: The signal envelope value E(t) reflects the overall amplitude characteristics of the signal and can show the changing trend and important characteristics of the signal; the square term of the original signal x(t) represents the square of the amplitude of the original signal. This term reflects the strength of the signal at time t and provides basic information about the signal amplitude. The instantaneous signal x h The square term of (t) represents the square of the amplitude of the instantaneous signal. This term reflects the instantaneous characteristics of the signal. In signal processing, the calculation of envelope values ​​usually utilizes the characteristics of complex signals. The original signal x(t) and its corresponding instantaneous signal x h (t) can be regarded as the real and imaginary parts of the complex signal, which means that the signal envelope value E(t) actually represents the amplitude of the complex signal.

[0034] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A fast and high-precision frequency measurement method for super-synchronous interharmonics, characterized in that: The method includes: Step S1: preset a reference fundamental wave signal, use a super-synchronous source to generate an initial fundamental wave signal on a discrete time series, use a nonlinear element to harmonically modulate the initial fundamental wave signal and output a modulated fundamental wave signal indicating that super-synchronization is completed; Step S2: setting the reference fundamental wave signal and the modulated fundamental wave signal in discrete time series, screening out the discrete signal components with abnormal frequencies in the modulated fundamental wave signal and marking them as abnormal components, and using a blind source separation algorithm to extract the abnormal frequency interval with frequency overlap in each abnormal component; Step S3: using an envelope detection method to extract abnormal fundamental wave data existing in all abnormal frequency intervals, and extracting all reference fundamental wave data in the reference fundamental wave signal; Step S4: Compare all abnormal fundamental wave data with the corresponding reference fundamental wave data at the same time series point, filter out the abnormal components of the abnormal fundamental wave data that are higher than the corresponding reference fundamental wave data, and perform precise frequency harmonic correction.

2. A fast and high-precision frequency measurement method for super-synchronous interharmonics according to claim 1, characterized in that: The blind source separation algorithm intercepts the abnormal frequency interval including: setting a power threshold for screening out abnormal frequency values, using the FastICA analysis method to extract the independent fundamental signal representing the existence of overlapping frequencies in the abnormal component in a manner that maximizes non-Gaussianity, performing FFT transformation on the independent fundamental signal and marking the output spectrum as a modulation spectrum, marking the frequency value corresponding to the power peak value in the modulation spectrum that is higher than the power threshold as the abnormal frequency value; and calculating the abnormal frequency interval representing the frequency overlapping area based on the half-peak width of the abnormal frequency value.

3. A fast and high-precision frequency measurement method for super-synchronous interharmonics according to claim 2, characterized in that: The abnormal frequency interval is calculated using a sliding window method based on the half-peak width of the abnormal frequency value. The process includes: setting the sliding window size based on the frequency resolution of the independent fundamental signal, initializing the sliding window from each abnormal frequency value to both ends of the frequency value axis, extracting the frequency components and corresponding power values ​​within the current window range at each sliding window position, and identifying the power peak within the window; When the power peak in the sliding window on both sides of the abnormal frequency value exceeds the power threshold, the sliding window slides one unit toward both ends of the frequency value axis; when the power peak in the sliding window on both sides of the abnormal frequency value is less than or equal to the power threshold, the frequency value corresponding to the power peak in the sliding window on both sides of the abnormal frequency value is set as the upper and lower boundary values ​​of the abnormal frequency interval.

4. A fast and high-precision frequency measurement method for super-synchronous interharmonics according to claim 3, characterized in that: The process of setting the sliding window size includes: assuming that the frequency resolution of the independent fundamental wave signal required by the sliding window is represented by Δf, assuming that the sampling frequency of the independent fundamental wave signal is represented by f s , let the number of FFT transformation points be N, the sliding window size be W, and set the adjustment constant k, Then the calculation formula of the sliding window W is expressed as: .

5. A fast and high-precision frequency measurement method for super-synchronous interharmonics according to claim 2, characterized in that: A white noise value for adding to FastICA analysis is set based on a modulation spectrum. The calculation and setting process of the white noise value includes: calculating the mean of the frequency values ​​in the modulation spectrum, and calculating the covariance matrix after subtracting the mean from each data point, performing eigenvalue decomposition on the covariance matrix, performing an inverse square root transformation, and then outputting the white noise value.

6. A fast and high-precision frequency measurement method for super-synchronous interharmonics according to claim 1, characterized in that: The specific contents of the envelope detection method include: extracting the signal envelope value of the abnormal component by using Hilbert transform; performing peak detection on the envelope signal to identify the significant peak in the envelope signal, after screening the corresponding time point on the envelope signal, extracting the corresponding abnormal fundamental wave data from the abnormal component according to the detected peak position.

7. A fast and high-precision frequency measurement method for super-synchronous interharmonics according to claim 6, characterized in that: The process of extracting the signal envelope value of the abnormal component by using Hilbert transform includes: obtaining the instantaneous signal of the abnormal component by using Hilbert transform, and calculating the signal envelope value based on the instantaneous signal.

8. A fast and high-precision frequency measurement method for super-synchronous interharmonics according to claim 7, characterized in that: The current time point of the design calculation is represented as t, and the integral variable is represented as Used to represent different time points in the original signal; The original signal of the abnormal component is expressed as , the instantaneous signal is represented by x h (t), Then the instantaneous signal x h The calculation formula of (t) is expressed as: .

9. A fast and high-precision frequency measurement method for super-synchronous interharmonics according to claim 8, characterized in that: Let the signal envelope value be E(t), and let the original signal at time t be x(t), Then the calculation formula of the signal envelope value E(t) is expressed as: .

10. A fast and high-precision frequency measurement method for super-synchronous interharmonics according to claim 7, characterized in that: Visualize the Hilbert transformed results and the extracted fundamental data, and apply a smoothing filter on the calculated signal envelope values.

Images