A method and device for high-precision measurement of harmonics and interharmonics of power broadband signals

Abstract

The application discloses a kind of electric power broadband signal harmonic / interharmonic high-precision measurement method and device.The method comprises: solving according to the construction of windowed signal spectrum asymptotic expression under constraint N linear equation group, design has analytic digital frequency interpolation coefficient and minimum sidelobe performance of normalized N order W window function, utilize different order W window to double weighting of broadband signal, carry out spectrum density detection based on amplitude spectrum, and measure broadband signal according to the difference value of detection spectrum line adaptive switching W window frequency interpolation factor and correction formula.The present application does not need to fit frequency interpolation coefficient, and has small amount of calculation, strong robustness, easy to realize, effectively realizes harmonic / interharmonic parameter real-time high-precision measurement in broadband range.

Description

Technical Field

[0001] This invention belongs to the field of power system broadband signal measurement technology, specifically relating to a method and apparatus for high-precision measurement of power broadband signal harmonics / interharmonics. Background Technology

[0002] With the increasing prevalence of power electronic devices in new power systems, the problem of dynamic interference in power signals is also growing. Harmonics and interharmonics have the most widespread impact, severely affecting the stable operation of the power grid and tasks such as power quality analysis, broadband oscillation monitoring, and early warning. Therefore, high-precision monitoring of broadband power signals in new power systems has become particularly important to ensure the efficient and safe operation of the power system. The "dual high" characteristics of new power systems have created new demands for signal measurement technologies.

[0003] Traditional methods for measuring harmonics / interharmonics in power signals can be categorized as follows: discrete frequency domain transformation methods, signal spatial spectrum estimation methods, least squares (RLS) state estimation methods, and adaptive parameter estimation methods based on artificial intelligence neural networks. Among these, traditional discrete frequency domain transformation methods suffer from picket fence effects and spectral leakage due to asynchronous sampling, significantly impacting the accuracy of signal parameter measurements. Furthermore, near-simultaneous spectral interference between frequencies makes them indistinguishable. Signal spatial spectrum estimation methods effectively reduce noise and offer high measurement accuracy, but their computational cost is too high, and their frequency resolution is insufficient, making real-time application in broadband measurement devices difficult. RLS methods require high precision in signal state space modeling and are prone to algorithm saturation. Artificial intelligence algorithms can perform precise signal analysis, but the accuracy of their parameter estimation depends on the accuracy of the initial tuning values ​​and the number of components; large frequency errors can lead to algorithm non-convergence.

[0004] Therefore, it is necessary to propose a new high-performance broadband signal harmonic / interharmonic measurement algorithm that balances computational efficiency, measurement accuracy, and numerical stability, and to design a corresponding broadband measurement device implementation framework. Summary of the Invention

[0005] Purpose of the invention: To address the shortcomings of existing technologies, this invention provides a method and apparatus for high-precision measurement of harmonics / interharmonics in broadband power signals, enabling efficient and accurate measurement of harmonic / interharmonic component parameters in broadband signals.

[0006] Technical solution: Firstly, a high-precision measurement method for harmonics / interharmonics of broadband power signals, comprising the following steps:

[0007] For N of a single frequency signal in the time domain sThe discrete Fourier transform is performed on the windowed weighted form to obtain its discrete spectrum. Based on the discrete spectrum of the windowed signal and the asymptotic expansion of the window function, the spectral ratio of adjacent spectral lines at the current frequency is calculated. The simplified form of the spectral ratio expression is used to obtain a polynomial for the digital frequency interpolation coefficients. With window function normalization as a condition and obtaining the analytical form of the digital frequency interpolation coefficients as the goal, an N-variable linear equation system is constructed to solve for the window function coefficients of the W window, obtaining the time-domain expression of the Nth-order W window. Using the bispectral analytical frequency interpolation method, the analytical forms of the amplitude correction and frequency interpolation coefficients corresponding to the Nth-order W window are obtained.

[0008] Based on the obtained Nth-order W-window, a broadband signal with a truncated window length of L is weighted using a low-order W-window with a relatively narrow main lobe and a high-order W-window with a relatively wide main lobe, respectively, to obtain two types of windowed signals. Discrete Fourier transforms are then performed on the original truncated signal and the two types of windowed signals to obtain discrete spectra. Frequency components are detected based on the amplitude spectrum of the original signal spectrum to obtain spectral lines at each frequency component. The frequency difference between each detected spectral line and its upper and lower spectral lines is calculated to obtain a frequency difference sequence over the entire spectrum. Based on the difference sequence of each spectral line, it is determined whether the difference of the currently detected spectral line is less than a given threshold. If so, the signal parameters are calculated using the discrete spectrum of the low-order W-window with a relatively narrow main lobe and the interpolation correction formula; otherwise, the discrete spectrum of the high-order W-window with a relatively wide main lobe and the correction formula are used for calculation.

[0009] Furthermore, based on the discrete spectrum of the windowed signal and the asymptotic expansion of the window function, the spectral ratio of adjacent spectral lines at the current frequency is calculated, including:

[0010] A cosine window is selected for window function design. Based on the discrete spectrum of the windowed signal and the asymptotic expansion of the trigonometric function of the cosine window, an approximate expression for the ratio of the spectra of adjacent spectral lines at the current frequency after weighting by the cosine window function is calculated:

[0011]

[0012] Among them, X λp (l) represents the form of the discrete spectrum at spectral line l, denoted by H(λ). p ,l,{a k}) represents the amplitude ratio between adjacent spectral lines in a discrete spectrum, λ p a is the digital frequency interpolation coefficient. k+1 These are the coefficients of the window function.

[0013] Furthermore, by simplifying the expression for the spectral line ratio, a polynomial for the digital frequency interpolation coefficients is obtained. Then, with window function normalization as a condition, and with the goal of obtaining the analytical form of the digital frequency interpolation coefficients, a system of N linear equations is constructed, including:

[0014] According to H(λ)p ,l,{a k}), which simplifies to the form of the following formula, thus utilizing H(λ) p ,l,{a k}) represents the digital frequency interpolation coefficient λ p :

[0015]

[0016] Regarding the digital frequency interpolation coefficient λ p polynomials Setting all coefficients of the terms where i>0 to 0, we obtain the following system of N linear equations:

[0017]

[0018] Where, β i,j This represents the weight corresponding to the j-th window function coefficient in the coefficient of the 2i-th term of the polynomial. Represents the term of degree 2i in a polynomial The corresponding completeness coefficient.

[0019] Furthermore, by solving for the window function coefficients of the W window, the time-domain expression of the Nth-order W window is obtained, including:

[0020] Find the unique solution for the window function coefficients of the W window in a system of N linear equations. N p Let represent the window length truncated by the Discrete Fourier Transform. A general form of the cosine window is chosen for designing the W window. The time-domain expression for the Nth-order cosine W window is:

[0021] Furthermore, the analytical forms of the amplitude correction and frequency interpolation coefficients corresponding to the Nth-order cosine W window are expressed as follows:

[0022]

[0023] Among them, A c (p) represents the amplitude correction corresponding to the Nth-order cosine W-window, λ p (N s ) represents the frequency interpolation coefficients corresponding to the Nth-order cosine W window.

[0024] Furthermore, the formulas required for calculating the Nth-order cosine W-window frequency interpolation correction and phase angle correction are expressed as follows:

[0025]

[0026] Where f p (N s ) represents the corrected digital frequency; Φ p (λ p) represents the corrected phase angle, angle represents the complex argument function, and f s Indicates the sampling frequency.

[0027] Secondly, a power broadband signal measuring device includes:

[0028] The synchronous sampling module performs synchronous sampling of signals based on clock data and analog-to-digital conversion to obtain digital signals, and filters out frequency components above the effective data frequency band of Nyquist sampling in the input signal through low-pass analog filtering.

[0029] The data storage unit stores dynamic waveform recording, continuous waveform recording, transient waveform recording data, as well as device configuration files and log files.

[0030] The broadband signal calculation module performs spectral analysis based on the effective signal obtained by synchronous sampling and low-pass filtering using the high-precision measurement method for power broadband signal harmonics / interharmonics described in the first aspect of the present invention, and obtains the parameter measurement values ​​of harmonics / interharmonics in the entire frequency band.

[0031] The data uploading module, according to the communication protocol, uses time-division multiplexing to upload the calculated harmonic component parameters of the previous whole second in phasor and frequency channels in order of harmonics from 1 to N. Based on the set frame number Fs, one set of harmonic parameters is uploaded per frame. For the calculated interharmonic component parameters, based on the set frame number, the interharmonics calculated at the whole second are sent in descending order of amplitude, with one set of interharmonic parameters uploaded per frame. Harmonic voltage and current amplitudes and frequencies are uploaded from the device phasor and analog A / B / C phase channels, respectively. Interharmonic voltage and current amplitudes and frequencies are also uploaded from the device phasor and analog A / B / C phase channels, respectively.

[0032] Thirdly, the present invention also provides a high-precision measurement system for harmonics / interharmonics of wideband power signals, comprising:

[0033] An N-order W-window design module for N-order time-domain single-frequency signals. s The discrete Fourier transform is performed on the windowed weighted form to obtain its discrete spectrum. Based on the discrete spectrum of the windowed signal and the asymptotic expansion of the window function, the spectral ratio of adjacent spectral lines at the current frequency is calculated. The simplified form of the spectral ratio expression is used to obtain a polynomial for the digital frequency interpolation coefficients. With window function normalization as a condition and obtaining the analytical form of the digital frequency interpolation coefficients as the goal, an N-variable linear equation system is constructed to solve for the window function coefficients of the W window, obtaining the time-domain expression of the Nth-order W window. Using the bispectral analytical frequency interpolation method, the analytical forms of the amplitude correction and frequency interpolation coefficients corresponding to the Nth-order W window are obtained.

[0034] The dual-weighted correction measurement module is used to weight a broadband signal with a truncated window length of L based on the obtained N-order W-window. This is achieved using a low-order W-window with a relatively narrow main lobe and a high-order W-window with a relatively wide main lobe, resulting in two types of windowed signals. Discrete Fourier transforms are then performed on both the original truncated signal and the two types of windowed signals to obtain discrete spectra. Frequency components are detected based on the amplitude spectrum of the original signal spectrum, yielding spectral lines at each frequency component. The frequency difference between each detected spectral line and its upper and lower spectral lines is calculated, resulting in a frequency difference sequence across the entire spectrum. Based on the difference sequence for each spectral line, it is determined whether the difference of the currently detected spectral line is less than a given threshold. If so, signal parameters are calculated using the discrete spectrum of the low-order W-window with a relatively narrow main lobe and the interpolation correction formula; otherwise, the discrete spectrum of the high-order W-window with a relatively wide main lobe and the correction formula are used for calculation.

[0035] Fourthly, the present invention provides a computer device comprising: one or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, wherein when the programs are executed by the processors, they implement the steps of the high-precision measurement method for power broadband signal harmonics / interharmonics as described in the first aspect.

[0036] Fifthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the high-precision measurement method for power broadband signal harmonics / interharmonics as described in the first aspect.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] (1) Based on asymptotic expansion and solving linear equations, an Nth-order W-window function with analytical digital frequency interpolation coefficients and the largest sidelobe attenuation rate is designed. This reduces the impact of spectral leakage after signal weighting, and analytical frequency interpolation avoids the complex polynomial calculations of coefficient fitting, reducing the time cost of frequency interpolation and improving the measurement efficiency of broadband signals. Since the cosine window interpolation form is simple, the invention uses a cosine window for W-window design. This construction method can also be extended to multinomial non-cosine windows (Gaussian windows, etc.) according to different requirements for interpolation efficiency and performance.

[0039] (2) Based on the main lobe width and side lobe attenuation properties, the signal is double-weighted by using a low-order W window with a narrow main lobe and a high-order W window with fast side lobe attenuation. The spectral difference is detected based on the original frequency domain transform amplitude spectrum. The frequency interpolation correction method based on different order W windows is adaptively switched for frequency components with different densities. This minimizes the impact of spectral leakage and spectral interference between components with similar frequencies, and achieves high-efficiency and accurate measurement of all harmonic / interharmonic parameters in a wide frequency range. Attached Figure Description

[0040] Figure 1 This is a flowchart of the broadband signal measurement method provided in this invention;

[0041] Figure 2 This is a partial waveform diagram of the broadband signal used in the simulation test of this invention;

[0042] Figure 3 This is a partial harmonic / interharmonic measurement spectrum diagram from the simulation test of this invention;

[0043] Figure 4 This is a comparison of the measurement errors of each component in the simulation test of this invention;

[0044] Figure 5 This is the system framework for implementing the broadband signal measurement device provided by the present invention. Detailed Implementation

[0045] The technical solutions in the embodiments of the present invention will now be clearly and completely described in conjunction with the accompanying drawings.

[0046] Example 1

[0047] This embodiment provides a high-precision measurement method for harmonics / interharmonics of broadband power signals. The main idea of ​​this method is as follows: Based on asymptotic expansion and solving linear equations, an Nth-order W-window function with analytical digital frequency interpolation coefficients and the largest sidelobe attenuation rate is designed. According to the main lobe width and sidelobe attenuation properties, different-order W-windows are used to double-weight the signal. The spectral difference is detected based on the amplitude spectrum of the original frequency domain transformation. The frequency interpolation correction method based on different-order W-windows is adaptively switched for frequency components with different densities, thereby accurately and efficiently measuring all harmonic / interharmonic parameters in a broadband range.

[0048] Reference Figure 1 The high-precision measurement method for harmonics / interharmonics of broadband power signals described in this invention specifically includes the following steps:

[0049] Step S1: Design an Nth-order W window with analytical frequency interpolation coefficients.

[0050] S1-1, First, the time-domain single-frequency signal u... p (t)=A p cos(2πΔf(l+λ p )k+φ p N s The discrete Fourier transform (DFT) with a truncated length of L is performed in a term-window weighted form, where N... sThe term "term number" refers to the number of terms in the general form of the window function (since the designed window function has not yet been introduced, it is used to indicate the number of terms). Physically, it represents the number of superimposed components in the window function, and in the proposed window function, it is equivalent to the order N of the window. The discrete spectrum of the windowed signal is obtained through DFT, and its general form at spectral line l is expressed as follows: Represented as:

[0051] Among them, A p ,φ p N represents the amplitude and phase angle of a single signal. p λ represents the length of the window truncated by the DFT. p a is the digital frequency interpolation coefficient. k+1 The coefficients of the window function are represented by Δf, which represents the frequency resolution. + (λ p ,k),f - (λ p ,k) represents the conjugate component in the discrete spectrum.

[0052] S1-2, Since the sampling window length L is a large integer, based on the discrete spectrum of the windowed signal and the asymptotic expansion of the trigonometric function of the cosine window, an approximate expression for the ratio of the spectra of adjacent spectral lines at the current frequency after weighting by the cosine window function is calculated:

[0053]

[0054] S1-3 uses the simplified form of the spectral line ratio expression to obtain a polynomial for the digital frequency interpolation coefficients, and constructs an N-variable linear equation system with the goal of obtaining the analytical form of the digital frequency interpolation coefficients, conditioned on window function normalization.

[0055] Based on the amplitude ratio H(λ) between adjacent spectral lines in the DFT spectrum p ,l,{a k}), which can be simplified to the form of the following formula, thus utilizing H(λ) p ,l,{a k}) represents the digital frequency interpolation coefficient λ p :

[0056]

[0057] Regarding the digital frequency interpolation coefficient λ p polynomials Setting all coefficients of the terms where i>0 to 0, we obtain the following system of N linear equations:

[0058]

[0059] Where, β i,jTable 2 shows the weights corresponding to the coefficients of the j-th window function in the coefficients of the 2i-th term of the polynomial. Represents the term of degree 2i in a polynomial The corresponding completeness coefficient.

[0060] S1-4, Solve for the unique solution of the window function coefficients of window W based on the above full-rank linear equation system. The time-domain expression of the Nth-order cosine W-window is obtained as follows:

[0061]

[0062] Table 1 shows the coefficients of the 3rd to 7th order window functions of the cosine W window obtained by constructing a system of linear equations. The order used in this embodiment of the invention is 3rd and 5th.

[0063] Table 1 lists the window function coefficients of the proposed Nth-order cosine W window.

[0064] 3rd order W window 3 / 8 1 / 2 1 / 8 4th order W window 5 / 16 15 / 32 3 / 16 1 / 32 5th-order W window 35 / 128 7 / 16 7 / 32 1 / 16 1 / 128 6th order W window 63 / 256 105 / 256 15 / 64 45 / 512 5 / 256 1 / 512 7th-order W window 231 / 1024 99 / 256 495 / 2048 55 / 512 33 / 1024 3 / 512 1 / 2048

[0065] The equations for solving the window function coefficients of the constructed 3rd and 5th order W windows are as follows:

[0066] • The general polynomial formula for cosine-type digital frequency interpolation coefficients:

[0067]

[0068] in N represents s The polynomial form corresponding to the term window function; λ p a represents the digital frequency interpolation coefficient. k This represents the coefficient of the k-th window function.

[0069] • Solving for the coefficients of the third-order window function:

[0070]

[0071] • Solving for the coefficients of the 5th-order window function:

[0072]

[0073] Where the equation coefficient β i,j By polynomial The part with zeros in the middle is obtained, and the coefficients of the cosine type are shown in Table 2 (the higher-order general formula is more complicated, and usually the calculation is performed directly after specifying the value of n). From this, a system of linear equations with unique solutions for the corresponding order window can be constructed, and the corresponding window function coefficients can be obtained.

[0074] Table 2. General formula for the coefficients of some equations in the form of a cosine window (order n+1).

[0075]

[0076]

[0077] S1-5, using the bispectral analytical frequency interpolation method, calculate the analytical forms of the amplitude correction, phase correction, and frequency interpolation coefficients corresponding to the Nth-order cosine W window, where the amplitude correction coefficient is expressed as:

[0078]

[0079] The analytical digital frequency interpolation coefficients are expressed as follows:

[0080]

[0081] By using the aforementioned frequency interpolation coefficients for interpolation, the non-analytical method of fitting the inverse function of the spectral ratio using a high-order polynomial to obtain the frequency interpolation coefficients, as required by traditional cosine window bispectral interpolation methods, is avoided, thus improving the computational efficiency of frequency interpolation. Furthermore, this method can be extended to multinomial non-cosine windows (such as Gaussian windows).

[0082] Step S2: Perform dual-weighted adaptive spectrum correction of the signal based on the N-order W window.

[0083] S2-1, using the W window obtained from the design, the signal is weighted by a window function;

[0084] Based on the main lobe width and side lobe attenuation properties, this invention employs different orders of W windows to perform dual weighting on the signal. Specifically, a low-order W window with a narrow main lobe (3rd order selected in this embodiment) and a high-order W window with a wide main lobe (5th order selected in this embodiment) are used to respectively weight the broadband signal U with a truncated window length of L. L (t) is weighted to obtain the windowed signal. and

[0085] S2-2, respectively, the original truncated signal U L (t), and windowing signal and Perform DFT calculations to obtain the discrete spectrum X L (k), and

[0086] S2-3, based on the original signal spectrum X L Frequency component detection is performed on the amplitude spectrum of (k) to obtain the spectral line l at the k-th frequency component. k k = 1, 2, ..., N h , where N h The number of harmonic / interharmonic components in a broadband signal;

[0087] S2-4, calculate the frequency difference between each detected spectral line and its upper and lower spectral lines to obtain the frequency difference sequence e(l) across the entire spectrum. k )=max{f(l k+i )-f(l k+i-1 )}, i=0,1, where f(l k ) indicates spectral line l k The frequency corresponding to i = 0 is l. k-1 ,l k The frequency difference between the two spectral lines, i=1 corresponds to l k ,l k+1 The frequency difference between the two spectral lines, e(l k ) indicates the current detected spectral line l k The maximum frequency difference between the two sides;

[0088] S2-5, based on the sequence of differences between the preceding and following spectral lines, set a threshold η to determine whether the difference of the currently detected k-th spectral line is less than the given threshold, i.e., determine e(l k If η < η, it indicates that the current frequency band has a dense frequency distribution and large spectral interference. The spectrum is then discretized using a third-order W-window with a narrow main lobe width. The signal parameters are calculated using the interpolation correction formula; otherwise, it indicates that the frequency distribution in the current frequency band is relatively dispersed and the spectral interference is small. A 5th-order W-window discrete spectrum with a wide main lobe width and a relatively fast side lobe attenuation rate is then used. Calculate using the correction formula;

[0089] Specifically, after threshold determination, based on the spectrum obtained by double-weighting the signal, the amplitude correction calculation formula is applied according to the combination of the 3rd-order W-window and the 5th-order W-window. Formula for calculating frequency interpolation coefficients Spectrum correction is performed, where the frequency correction calculation formula is: The formula for phase angle correction is as follows: Where f p (N s ) represents the corrected digital frequency; Φ p (λ p ) represents the corrected phase angle, angle represents the complex argument function, and f s Indicates the sampling frequency.

[0090] Step S2-6: Determine if k is the last detected frequency component. If so, end the loop and output all component parameters; otherwise, continue with step S2-3 until all frequency components have been detected.

[0091] To verify the performance of the proposed method, simulation tests were performed in the embodiments. Figure 2The waveform diagram is a partial waveform of the broadband signal used for simulation testing. The broadband test signal is superimposed with 40 components, including 1 fundamental frequency and 39 harmonics / interharmonics. The frequency intervals between the components other than the fundamental frequency are randomly switched at 1.5Hz, 5Hz and 100Hz, and the phase angle and amplitude are randomly assigned to improve the measurement performance of the simulation algorithm under different signal conditions.

[0092] Figure 3 The image shows the partial harmonic / interharmonic measurement spectrum obtained from the simulation test using the method of this invention. A sliding window sampling calculation was performed using a sampling frequency of Fs = 12.8 kHz and a window length of L = 2^15, with each sliding window acquiring 2.56 seconds of data. It can be seen that the method proposed in this invention overcomes the amplitude averaging effect caused by the original signal's spectral leakage, as well as the impact of signal windowing on the resolution of similar frequencies. Through a dual-weighted adaptive switching spectrum correction method, the parameters of each component in the original signal are accurately estimated.

[0093] Figure 4 To compare the measurement errors of each component in the simulation test, the experiment used a four-term 5th-order Nuttall window interpolation FFT algorithm with high spectral correction accuracy for measurement performance comparison. The measurement results show that in frequency bands where the components are far apart, both the proposed method and the comparative method have high measurement accuracy due to the absence of spectral interference, but the proposed method's accuracy is still higher than the comparative method. In frequency bands where components are densely distributed, the proposed method's measurement accuracy is much higher than the comparative method. This is because although the four-term 5th-order Nuttall window has a fast sidelobe attenuation rate and can effectively suppress spectral leakage, its amplitude-frequency main lobe is too wide, resulting in reduced resolution of close frequencies after windowing. The proposed method, however, can switch between the narrower 3rd-order W window and the wider 5th-order W window through dual weighting and spectral difference determination, achieving adaptive measurement of broadband components.

[0094] Table 3 shows the harmonic / interharmonic setpoints and corresponding measurement errors used in the simulation test for the proposed method. The amplitude error is represented by the proportion of the measured harmonic / interharmonic amplitude, while the phase and frequency measurement errors are expressed in the original units of the measured data. According to the measurement results, the maximum frequency error is 0.00338 Hz, the maximum amplitude error as a percentage of the current component amplitude is 0.288%, and the maximum phase angle error is 1.556 degrees, all of which are lower than relevant industry standards.

[0095] Table 3. Set values ​​of broadband signal harmonic / interharmonic parameters and measurement errors of the proposed method.

[0096]

[0097]

[0098] Example 2

[0099] Reference Figure 5 This embodiment provides a power broadband signal measurement device, including:

[0100] The synchronous sampling module performs synchronous sampling of signals based on clock data and ADC (analog-to-digital converter) to obtain digital signals. It also filters out frequency components above the effective data frequency band of Nyquist sampling in the input signal through preprocessing methods such as low-pass analog filtering to prevent frequency aliasing.

[0101] The data storage unit (not shown in the figure) stores dynamic waveform recording, continuous waveform recording, transient waveform recording data, as well as device configuration files, log files, etc.

[0102] The broadband signal calculation module performs spectral analysis based on the effective signal obtained from synchronous sampling and data preprocessing, using the high-precision measurement method for harmonics / interharmonics of broadband power signals described in Example 1, to obtain the parameter measurement values ​​of harmonics / interharmonics across the entire frequency band.

[0103] The data uploading module, in accordance with the GBT 26865.2-2011 communication protocol, uses time-division multiplexing to upload the calculated harmonic component parameters of the previous whole second in phasor and frequency channels according to the order of harmonics from 1 to N. According to the set frame number Fs, a set of harmonic parameters is uploaded for each frame. For the calculated interharmonic component parameters, according to the set frame number, the interharmonics calculated at the whole second are sent in descending order of amplitude, and a set of interharmonic parameters is uploaded for each frame. The three-phase harmonics, interharmonic voltage, current amplitude, and frequency are uploaded from the A\B\C phase channels of the device phasor and the A\B\C phase channels of the analog quantity, respectively.

[0104] Example 3

[0105] This embodiment provides a high-precision measurement system for harmonics / interharmonics of wideband power signals, including:

[0106] An N-order W-window design module for N-order time-domain single-frequency signals. s The discrete Fourier transform is performed on the windowed weighted form to obtain its discrete spectrum. Based on the discrete spectrum of the windowed signal and the asymptotic expansion of the window function, the spectral ratio of adjacent spectral lines at the current frequency is calculated. The simplified form of the spectral ratio expression is used to obtain a polynomial for the digital frequency interpolation coefficients. With window function normalization as a condition and obtaining the analytical form of the digital frequency interpolation coefficients as the goal, an N-variable linear equation system is constructed to solve for the window function coefficients of the W window, obtaining the time-domain expression of the Nth-order W window. Using the bispectral analytical frequency interpolation method, the analytical forms of the amplitude correction and frequency interpolation coefficients corresponding to the Nth-order W window are obtained.

[0107] The dual-weighted correction measurement module is used to weight a broadband signal with a truncated window length of L based on the obtained N-order W-window. This is achieved using a low-order W-window with a relatively narrow main lobe and a high-order W-window with a relatively wide main lobe, resulting in two types of windowed signals. Discrete Fourier transforms are then performed on both the original truncated signal and the two types of windowed signals to obtain discrete spectra. Frequency components are detected based on the amplitude spectrum of the original signal spectrum, yielding spectral lines at each frequency component. The frequency difference between each detected spectral line and its upper and lower spectral lines is calculated, resulting in a frequency difference sequence across the entire spectrum. Based on the difference sequence for each spectral line, it is determined whether the difference of the currently detected spectral line is less than a given threshold. If so, signal parameters are calculated using the discrete spectrum of the low-order W-window with a relatively narrow main lobe and the interpolation correction formula; otherwise, the discrete spectrum of the high-order W-window with a relatively wide main lobe and the correction formula are used for calculation.

[0108] It should be understood that the high-precision measurement system for power broadband signal harmonics / interharmonics provided in this embodiment can realize all the technical solutions in the above method embodiments. The functions of each functional module can be specifically implemented according to the methods in the above method embodiments. The specific implementation process can be referred to the relevant descriptions in the above embodiments, which will not be repeated here.

[0109] Example 4

[0110] This embodiment provides a computer device, the device including one or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of a high-precision measurement method for harmonics / interharmonics of power broadband signals as described in Embodiment 1.

[0111] Example 5

[0112] This embodiment provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the steps of a high-precision measurement method for harmonics / interharmonics of a power broadband signal as described in Embodiment 1.

[0113] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus (systems), computer devices, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied 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.

[0114] This invention is described with reference to a flowchart of a method according to embodiments of the invention. It should be understood that each step in the flowchart and combinations thereof can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing device, generate instructions for implementing the process. Figure 1 A device for a function specified in one or more processes.

[0115] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 The function specified in one or more processes.

[0116] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 Steps of a specified function in one or more processes.

Claims

1. A high-precision measurement method for harmonics / interharmonics of wideband power signals, characterized in that, Includes the following steps: For N of a single frequency signal in the time domain s The discrete Fourier transform is performed in the windowed weighted form to obtain its discrete spectrum. Based on the discrete spectrum of the windowed signal and the asymptotic expansion of the window function, the spectral ratio of adjacent spectral lines at the current frequency is calculated. By simplifying the expression for the spectral line ratio, a polynomial for the digital frequency interpolation coefficients is obtained. With window function normalization as a condition and the goal of obtaining the analytical form of the digital frequency interpolation coefficients, an N-variable linear equation system is constructed. The window function coefficients of the W window are solved to obtain the time-domain expression of the Nth-order W window. Using the bispectral line analytical frequency interpolation method, the analytical forms of the amplitude correction and frequency interpolation coefficients corresponding to the Nth-order W window are obtained. Based on the obtained Nth-order W window, the broadband signal with a truncated window length of L is weighted by the low-order W window with a relatively narrow main lobe and the high-order W window with a relatively wide main lobe, respectively, to obtain two types of windowed signals. Discrete Fourier transform is then performed on the original truncated signal and the two types of windowed signals to obtain discrete spectra. Frequency components are detected based on the amplitude spectrum of the original signal spectrum to obtain the spectral lines at each frequency component. Calculate the frequency difference between each detected spectral line and its upper and lower spectral lines to obtain a frequency difference sequence across the entire spectrum. Based on the difference sequence of each spectral line, determine whether the difference of the currently detected spectral line is less than a given threshold. If so, use a low-order W-window discrete spectrum with a relatively narrow main lobe width and an interpolation correction formula to calculate the signal parameters. Otherwise, use a high-order W-window discrete spectrum with a relatively wide main lobe width and a correction formula for calculation. The calculation of the spectral ratio of adjacent spectral lines at the current frequency is based on the discrete spectrum of the windowed signal and the asymptotic expansion of the window function, including: A cosine window is selected for window function design. Based on the discrete spectrum of the windowed signal and the asymptotic expansion of the trigonometric function of the cosine window, an approximate expression for the ratio of the spectra of adjacent spectral lines at the current frequency after weighting by the cosine window function is calculated: in, Representing the form of the discrete spectrum at spectral line l, in This represents the amplitude ratio between adjacent spectral lines in a discrete spectrum. These are digital frequency interpolation coefficients. This represents the coefficient of the k-th window function; By simplifying the expression for spectral line ratios, a polynomial for the digital frequency interpolation coefficients is obtained. Then, with window function normalization as a condition, and aiming to obtain the analytical form of the digital frequency interpolation coefficients, a system of N linear equations is constructed, including: according to Simplify it to the form of the following formula, and then utilize... To represent digital frequency interpolation coefficients : Regarding digital frequency interpolation coefficients polynomials Setting all coefficients of the terms where i>0 to 0, we obtain the following system of N linear equations: in, This represents the weight corresponding to the j-th window function coefficient in the coefficient of the 2i-th term of the polynomial. Represents the term of degree 2i in a polynomial The corresponding completeness coefficient.

2. The method according to claim 1, characterized in that, Solving for the window function coefficients of the W window yields the time-domain expression of the Nth-order W window, including: Find the unique solution for the window function coefficients of the W window in a system of N linear equations. , Let represent the window length truncated by the Discrete Fourier Transform. A general form of the cosine window is chosen for designing the W window. The time-domain expression for the Nth-order cosine W window is: .

3. The method according to claim 2, characterized in that, The analytical forms of the amplitude correction and frequency interpolation coefficients corresponding to the Nth-order cosine W window are expressed as follows: in, For amplitude correction corresponding to the Nth order cosine W window, These are the frequency interpolation coefficients corresponding to the Nth-order cosine W-window.

4. The method according to claim 3, characterized in that, The formulas required for calculating Nth-order cosine W-window frequency interpolation correction and phase angle correction are expressed as follows: in Indicates the corrected digital frequency; This represents the corrected phase angle, where angle represents the complex argument function. Indicates the sampling frequency.

5. A power broadband signal measuring device, characterized in that, include: The synchronous sampling module performs synchronous sampling of signals based on clock data and analog-to-digital conversion to obtain digital signals, and filters out frequency components above the effective data frequency band of Nyquist sampling in the input signal through low-pass analog filtering. The data storage unit stores dynamic waveform recording, continuous waveform recording, transient waveform recording data, as well as device configuration files and log files. The broadband signal calculation module performs spectral analysis based on the effective signal obtained by synchronous sampling and low-pass filtering using the high-precision measurement method for harmonics / interharmonics of broadband power signals as described in any one of claims 1-4, to obtain the parameter measurement values ​​of harmonics / interharmonics across the entire frequency band. The data uploading module, according to the communication protocol, uses time-division multiplexing to upload the calculated harmonic component parameters of the previous whole second in phasor and frequency channels in order of harmonics from 1 to N. Based on the set frame number Fs, one set of harmonic parameters is uploaded per frame. For the calculated interharmonic component parameters, based on the set frame number, the interharmonics calculated at the whole second are sent in descending order of amplitude, with one set of interharmonic parameters uploaded per frame. Harmonic voltage and current amplitudes and frequencies are uploaded from the device phasor and analog A / B / C phase channels, respectively. Interharmonic voltage and current amplitudes and frequencies are also uploaded from the device phasor and analog A / B / C phase channels, respectively.

6. A high-precision measurement system for harmonics / interharmonics of wideband power signals, used to implement the method described in any one of claims 1-4, characterized in that, include: An N-order W-window design module for N-order time-domain single-frequency signals. s The discrete Fourier transform is performed in the windowed weighted form to obtain its discrete spectrum. Based on the discrete spectrum of the windowed signal and the asymptotic expansion of the window function, the spectral ratio of adjacent spectral lines at the current frequency is calculated. By simplifying the expression for the spectral line ratio, a polynomial for the digital frequency interpolation coefficients is obtained. With window function normalization as a condition and the goal of obtaining the analytical form of the digital frequency interpolation coefficients, an N-variable linear equation system is constructed. The window function coefficients of the W window are solved to obtain the time-domain expression of the Nth-order W window. Using the bispectral line analytical frequency interpolation method, the analytical forms of the amplitude correction and frequency interpolation coefficients corresponding to the Nth-order W window are obtained. The dual-weighted correction measurement module is used to weight a broadband signal with a truncated window length of L based on the obtained N-order W window, using a low-order W window with a relatively narrow main lobe and a high-order W window with a relatively wide main lobe, respectively, to obtain two types of windowed signals. Discrete Fourier transforms are then performed on the original truncated signal and the two types of windowed signals to obtain discrete spectra. Frequency components are detected based on the amplitude spectrum of the original signal spectrum to obtain spectral lines at each frequency component. Calculate the frequency difference between each detected spectral line and its upper and lower spectral lines to obtain a frequency difference sequence across the entire spectrum. Based on the difference sequence of each spectral line, determine whether the difference of the currently detected spectral line is less than a given threshold. If so, use a low-order W-window discrete spectrum with a relatively narrow main lobe width and an interpolation correction formula to calculate the signal parameters. Otherwise, use a high-order W-window discrete spectrum with a relatively wide main lobe width and a correction formula for calculation.

7. A computer device, characterized in that, include: One or more processors; Memory; And one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, wherein when the programs are executed by the processors, they implement the steps of the high-precision measurement method for power broadband signal harmonics / interharmonics as claimed in any one of claims 1-4.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the high-precision measurement method for harmonics / interharmonics of power broadband signals as described in any one of claims 1-4.

Images