A continuous segmented quadratic identification multi-frequency signal fast frequency measurement method

Abstract

The application discloses a kind of continuous segmented secondary identification multi-frequency signal fast frequency measurement method, can realize the frequency fast and accurate measurement of wideband multi-frequency signal.First, according to signal frequency range, select appropriate resolution, after resampling, the original sampling signal is carried out primary FFT transformation, obtains the approximate range of signal frequency distribution;Then according to the resolution, the entire signal band is divided into uniform multiple, and the corresponding band-pass digital filter is designed for each section.The result after primary FFT is mapped on segmented band, the frequency band where signal frequency is located is selected, and after filtering processing, complex modulation is carried out to refine spectrum analysis, so as to complete the accurate measurement of wideband multi-frequency signal frequency.The application selects the frequency band where signal is located on the basis of primary FFT, and further refines secondary identification, can reduce the time consumption of embedded device for full-band signal scanning, improve the frequency resolution of signal analysis, realize the fast and accurate measurement of frequency.

Description

Technical Field

[0001] This invention belongs to the field of power automation technology, specifically relating to a rapid frequency measurement method for multi-frequency signals with continuous segmented secondary identification. Background Technology

[0002] With the development and popularization of renewable energy sources such as photovoltaics and wind power, and the continuous development of smart grids, power grids are showing a trend towards complex interconnection and high levels of power electronics. This power electronics-ization of the grid has highlighted power quality issues and introduced a large number of broadband signals, including interharmonics and higher harmonics, into the grid. Phenomena such as electromagnetic oscillations, low-frequency oscillations, and sub / supersynchronous oscillations are also increasing in the grid. Real-time and accurate measurement of the components of the grid signals is a crucial guarantee for the safe and stable operation of the power grid.

[0003] In the measurement of harmonic and interharmonic signals, the Fast Fourier Transform (FFT) is the most widely used spectral analysis method. However, to measure interharmonic signals with frequency intervals of 1 Hz or oscillation signals on the order of 0.1 Hz over a wide frequency range, a sufficiently large number of sampling points is required, inevitably leading to a large computational load on the equipment. This makes it difficult to meet the needs of broadband signal measurement applications in the context of power electronics. Therefore, a new technical solution is needed to address the aforementioned technical problems. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a rapid frequency measurement method for multi-frequency signals with continuous segmentation and secondary identification. This method solves the frequency aliasing phenomenon that occurs when the difference between multiple frequency points in the sampled signal is less than Δf, reduces the time consumed by embedded devices to scan the full frequency band signal, improves the frequency resolution of signal analysis, and enables rapid and accurate frequency measurement.

[0005] The specific technical solution adopted in this invention is as follows:

[0006] In a first aspect, the present invention provides a method for rapid frequency measurement of multi-frequency signals with continuous segmented secondary identification, comprising:

[0007] S1, acquire wideband raw sampling data, based on the preset maximum frequency F. max The original broadband sampling data is low-pass filtered to obtain the low-pass filtered data.

[0008] S2, based on the sampling rate f of the broadband raw sampling data s and the number of sampling points N s Select the frequency resolution Δf for the initial scan;

[0009] S3. For the low-pass filtered data, extract N sampling points according to the frequency resolution Δf, and perform an N-point Fast Fourier Transform (FFT) operation to obtain the spectrum of the initial scan, where N = f. s / Δf;

[0010] S4, the frequency band of the initial scan is divided into N-1 frequency intervals F according to the frequency resolution Δf. b ′ n For n∈(0,N-1), the spectrum of the initial scan is mapped to the frequency interval to obtain the mapped frequency interval F. b ′ m , m∈(0,N-1);

[0011] S5, based on the mapped frequency interval F b ′ m Design the passband range of a bandpass digital filter [f] a ,f b The designed bandpass digital filter is used to perform bandpass filtering on the original wideband sampling data to obtain the bandpass filtered data.

[0012] S6, Perform complex modulation refinement spectrum analysis on the bandpass filtered data to obtain the bandpass frequency range [f a ,f b The frequency of the data;

[0013] S7, responding to the bandpass frequency range [f] a ,f b Exceeding F max Complete full-band signal frequency measurement; responds to the bandpass interval [f a ,f b Not exceeding F max Return to S4 and proceed to the next frequency range processing flow.

[0014] In some embodiments, S1: the low-pass filtering process employs an FIR digital low-frequency filter, with the cutoff frequency set to F. c =F max .

[0015] In some embodiments, S2, the frequency resolution Δf ≥ f s / N s The Δf value can be adjusted according to the actual system.

[0016] In some embodiments, S4 includes:

[0017] For the frequency band F of the initial scan b =[0,F max The frequency range is divided into N-1 frequency ranges F based on the frequency resolution Δf. b ′ n =[n*Δf,(n+2)*Δf],n∈(0,N-1);

[0018] Map the spectrum of the initial scan to a frequency range: The spectrum of the initial scan has a frequency extremum at m*Δf, m∈(0,N-1), which is mapped to the frequency range F. b ′ m =[(m-1)*Δf,(m+1)*Δf].

[0019] In some embodiments, S5, F b ′ m =[(m-1)*Δf,(m+1)*Δf], f a = (m-1)*Δf,f b = (m+1)*Δf, where m represents the number of points corresponding to the spectrum currently being analyzed by FFT.

[0020] In some embodiments, in S5, a bandpass digital filter is designed using the equal ripple optimal approximation method, and the bandpass digital filter is a finite-length unit impulse response (FIR) digital filter.

[0021] FIR digital filters can be made with strictly linear phase and arbitrary amplitude characteristics, and are always stable. FIR filter design mainly employs the window function method, frequency sampling method, and equal-ripple method. The equal-ripple optimal approximation method is an optimal design method; filters designed using this method have the smallest maximum error in frequency response relative to an ideal filter. The time delay of bandpass digital filters is mainly affected by the order and cutoff frequency, using H... d (ω) represents the amplitude characteristic function that we want to approximate. When designing a linear-phase FIR digital filter, H d (ω) must satisfy the linear phase constraint. Let H(ω) denote the amplitude characteristic function of the actual designed filter. Define the weighted error function ε(ω) as follows:

[0022] ε(ω)=W(ω)[H d (ω)-H(ω)]

[0023] In the formula, W(ω) is the amplitude error weighting function, used to control the amplitude approximation accuracy of different frequency bands. The equiripple optimal approximation method calculates the filter coefficient vector h(n) to minimize the maximum absolute amplitude error |ε(ω)| in both the passband and stopband.

[0024] When designing a bandpass filter, the bandpass frequency follows fpass. a and f b The settings, including the boundary frequency, maximum passband attenuation, and maximum stopband attenuation coefficient, should be determined to minimize the filter order n within the error requirements calculated by the device.

[0025] In some embodiments, in S6, complex modulation refined spectrum analysis ZoomFFT (ZoomFFT) is used to perform fine analysis on the data after bandpass filtering. The sampling interval after resampling is D, and the frequency resolution becomes Δf1=Δf / D, where D is the decimation ratio. The N-point time-domain sequence obtained after resampling is subjected to fast Fourier transform, and the resolution is improved by a factor of D.

[0026] In a second aspect, the present invention provides a fast frequency measurement device for multi-frequency signals with continuous segmented secondary identification, including a processor and a storage medium;

[0027] The storage medium is used to store instructions;

[0028] The processor is configured to operate according to the instructions to perform the steps of the method according to the first aspect.

[0029] Thirdly, the present invention provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect.

[0030] The beneficial effects achieved by the present invention, a rapid frequency measurement method for multi-frequency signals with continuous segmentation and secondary identification, are as follows: To meet the requirements of adaptive measurement algorithms for complex and wideband power grid signals, and to solve the frequency aliasing phenomenon that occurs when the difference between multiple frequency points in the sampled signal is less than Δf, a rapid frequency measurement method for multi-frequency signals with continuous segmentation and secondary identification is provided. This reduces the time consumed by embedded devices to scan the full-band signal, improves the frequency resolution of signal analysis, and achieves rapid and accurate frequency measurement. Attached Figure Description

[0031] Figure 1 This is a flowchart of a rapid frequency measurement method for multi-frequency signals with continuous segmentation and secondary identification according to an embodiment of the present invention;

[0032] Figure 2 This is a spectrum secondary scan mapping interval diagram in the embodiment. Detailed Implementation

[0033] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0034] In the description of this invention, the terms "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0035] Example 1

[0036] A fast frequency measurement method for multi-frequency signals with continuous segmentation and secondary identification includes:

[0037] S1, acquire wideband raw sampling data, based on the preset maximum frequency F. max The original broadband sampling data is low-pass filtered to obtain the low-pass filtered data.

[0038] S2, based on the sampling rate f of the broadband raw sampling data s and the number of sampling points N s Select the frequency resolution Δf for the initial scan;

[0039] S3. For the low-pass filtered data, extract N sampling points according to the frequency resolution Δf, and perform an N-point Fast Fourier Transform (FFT) operation to obtain the spectrum of the initial scan, where N = f. s / Δf;

[0040] S4, the frequency band of the initial scan is divided into N-1 frequency intervals F according to the frequency resolution Δf. b ′ n For n∈(0,N-1), the spectrum of the initial scan is mapped to the frequency interval to obtain the mapped frequency interval F. b ′ m , m∈(0,N-1);

[0041] S5, based on the mapped frequency interval F b ′ m Design the passband range of a bandpass digital filter [f] a ,f b The designed bandpass digital filter is used to perform bandpass filtering on the original wideband sampling data to obtain the bandpass filtered data.

[0042] S6, Perform complex modulation refinement spectrum analysis on the bandpass filtered data to obtain the bandpass frequency range [f a ,f b The frequency of the data;

[0043] S7, responding to the bandpass frequency range [f] a ,f b Exceeding F max Complete full-band signal frequency measurement; responds to the bandpass interval [f a ,f b Not exceeding F max Return to S4 and proceed to the next frequency range processing flow.

[0044] In some embodiments, reference Figure 1 As shown in the figure, this embodiment proposes a fast frequency measurement method for multi-frequency signals with continuous segmentation and secondary identification. The method includes the following steps:

[0045] Step 1: Set the maximum calculable frequency to F according to application requirements and the A / D sampling rate of the broadband measurement device. max The raw sampling data from the wideband measurement device's A / D converter is low-pass filtered. The low-pass filter is designed using an FIR type filter with a filter cutoff frequency F. c =F max Based on application requirements, F is currently involved. max =2.5kHz.

[0046] Step 2, based on the sampling rate f of the original A / D sampling data from the broadband measurement device. s and the number of sampling points N s Select and set an appropriate initial frequency scan resolution Δf (Δf ≥ f s / N s The value of Δf can be adjusted according to the actual system. The Fast Fourier Transform (FFT) first requires truncating and discretizing the infinitely long continuous signal, and then performing a FFT on the resulting finite-length signal sequence with a fixed sampling interval. This truncation of the continuous time-domain signal is equivalent to adding a rectangular window, where f is set... s =6400Hz, N s =4096, the frequency resolution Δf>f can be appropriately increased. s / N s This reduces the computational cost of FFT.

[0047] Step 3: Extract N sampling points from the low-pass filtered data according to the set frequency resolution Δf, where N = f s / Δf, perform an N-point FFT operation to obtain the spectrum of the initial scan, where When using a rectangular window function to analyze continuous time-domain signals, the truncation causes discontinuities at the boundaries, dispersing energy concentrated at a particular frequency into nearby frequency domains, leading to erroneous spectral analysis results. Methods to reduce spectral analysis errors caused by this problem include increasing the window length or using a windowing interpolation function to ensure the signal decays slowly and smoothly at the truncation point and approaches zero at the boundaries, thus reducing spectral leakage. An ideal window function has a narrow main lobe and small side lobes, effectively capturing the target spectral components while minimizing their mutual interference.

[0048] Based on the requirements for measuring electrical signals, after comparing and selecting parameters of all commonly used window functions, several suitable window functions were selected. The characteristic diagrams of various window functions such as Hamming window, Hanning window, and Blackman window are shown in Table 1.

[0049] Table 1. Window functions and their characteristic parameters

[0050]

[0051] Step 4, measure the signal frequency band F b =[0,F max The frequency interval is divided according to the frequency resolution Δf, and the resulting interval F b ′ n = [n*Δf, (n+2)*Δf], n∈(0,N-1). Map the spectrum after the initial FFT scan to the frequency range; for example... Figure 2 As shown, if there exists a frequency extremum at m*Δf, m∈(0,N-1), then it is mapped to F b ′ m = [(m-1)*Δf, (m+1)*Δf] interval.

[0052] Step 5: Map the frequency range F of the extreme value spectrum. b ′ m Design a bandpass digital filter with a passband range of [f]. a ,f b ], where f a = (m-1)*Δf,f b = (m+1)*Δf, where m represents the number of points corresponding to the spectrum currently being analyzed by FFT;

[0053] Finite Impulse Response (FIR) digital filters are employed. FIR digital filters can be designed with strictly linear phase, arbitrary amplitude characteristics, and are always stable. Generally, higher filter orders result in greater stopband attenuation, smaller passband error, narrower transition band, and longer delay. Therefore, it is necessary to design an optimal filter to achieve the best filtering performance at the lowest possible order. FIR filter design primarily employs the window function method, frequency sampling method, and equal-ripple method. The disadvantages of the window function method are: it is difficult to design a filter with a pre-defined cutoff frequency; and the resulting filter order is usually too high to meet the same design specifications. The disadvantage of the frequency sampling method is the limitation on the cutoff frequency value. Furthermore, the approximation errors of the window function method and the frequency sampling method are not uniformly distributed across the frequency band; the error is larger near the band edge and smaller further away. The equal-ripple optimal approximation method is an optimal design method; filters designed using this method have the smallest maximum error in frequency response relative to the ideal filter.

[0054] The time delay of a bandpass digital filter is mainly affected by its order and cutoff frequency, denoted by H. d (ω) represents the amplitude characteristic function that we want to approximate. When designing a linear-phase FIR digital filter, G d (ω) must satisfy the linear phase constraint. Let H(ω) denote the amplitude characteristic function of the actual designed filter. Define the weighted error function ε(ω) as follows:

[0055] ε(ω)=W(ω)[H d (ω)-H(ω)]

[0056] In the formula, W(ω) is the amplitude error weighting function, used to control the amplitude approximation accuracy of different frequency bands. The equiripple optimal approximation method calculates the filter coefficient vector h(n) to minimize the maximum absolute amplitude error |ε(ω)| in both the passband and stopband.

[0057] When designing a bandpass filter, the bandpass frequency follows fpass. a and f b The settings, including the boundary frequency, maximum passband attenuation, and maximum stopband attenuation coefficient, should be determined to minimize the filter order n within the error requirements calculated by the device.

[0058] Step 6: Perform complex modulation refinement spectrum analysis (ZoomFFT) on the data after bandpass filtering.

[0059] Complex modulation refinement spectral analysis is a high-frequency resolution Fourier transform method based on signal complex modulation, abbreviated as ZoomFFT. It can refine and amplify a specific frequency domain of a signal, significantly improving the frequency resolution of that band. Currently, the frequency shifting method is mainly used, and the specific steps are as follows:

[0060] (1) After sampling the signal y(t), the sampling sequence y0(n) is obtained. At this time, the sampling length is the product of the thinning factor D and the length N of the ordinary FFT, i.e. DN;

[0061] (2) Perform complex modulation frequency shift on y0(n), i.e., multiply by the unit rotation factor e. -j2πnfmid / fs This moves the frequency origin from zero to the center f of the frequency band to be refined (f1, f2). mid f mid = (f2+f1) / 2, which is equivalent to moving the center of the refined frequency band to zero, forming a frequency band with f mid Let y(n) be a new signal at the zero frequency point.

[0062] (3) To avoid spectral aliasing during subsequent resampling, y(n) is subjected to low-pass digital filtering to remove components outside the frequency band that need to be refined. Let the sampling interval for resampling be D, then the sampling frequency becomes f. s / D, where D is the selection-to-decimation ratio. According to the sampling theorem, the cutoff frequency of the low-pass filter is f. s / 2D;

[0063] (4) After performing an inverse Fourier transform of length DN on the filtered data to obtain the time-domain sequence, resampling is performed, and the sampling frequency of the resampling becomes f. s / D;

[0064] (5) Perform a Fast Fourier Transform on the N-point time-domain sequence obtained after resampling, at which point the resolution f s Increase to (f s / D) / N=f s / DN increases the resolution by a factor of D, hence D is also called the refinement factor.

[0065] (6) Move the FFT calculation results to the actual frequency. Since all data are useful information when FFT calculation is performed after resampling, it is necessary to move the calculated negative frequency components to the correct position.

[0066] After the above steps, the spectral characteristics of the original signal sequence within a specific frequency range can be effectively obtained, achieving refinement and amplification, and the frequency resolution is improved by D times.

[0067] Step 7, complete the bandpass interval [f] a ,f bFrequency analysis within the specified range will proceed to the next frequency band range. If the frequency range exceeds F... max Then the analysis ends.

[0068] Example 2

[0069] Secondly, this embodiment provides a fast frequency measurement device for multi-frequency signals with continuous segmentation and secondary identification, including a processor and a storage medium;

[0070] The storage medium is used to store instructions;

[0071] The processor is configured to operate according to the instructions to perform the steps of the method according to Embodiment 1.

[0072] Example 3

[0073] Thirdly, this embodiment provides a storage medium on which a computer program is stored, which, when executed by a processor, implements the steps of the method described in Embodiment 1.

[0074] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application 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.

[0075] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, 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 apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0076] 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 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0077] 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 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0078] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A rapid frequency measurement method for multi-frequency signals with continuous segmentation and secondary identification, characterized in that, include: S1, acquire wideband raw sampling data, based on a preset maximum frequency. The original broadband sampling data is low-pass filtered to obtain the low-pass filtered data. S2, based on the sampling rate of the broadband raw sampling data and number of sampling points Select the frequency resolution for the initial scan. ; S3, adjusts the low-pass filtered data according to frequency resolution. N sampling points are extracted, and an N-point Fast Fourier Transform (FFT) operation is performed to obtain the spectrum of the initial scan. ; S4, for the frequency band of the initial scan. According to frequency resolution Divide the frequency range into sections. Frequency range Mapping the spectrum of the initial scan to frequency ranges: The spectrum responding to the initial scan in... , There exists a frequency extremum at a certain point, which is mapped to a frequency range. ; S5, based on the mapped frequency range Design the bandpass frequency range of a bandpass digital filter. The designed bandpass digital filter is used to perform bandpass filtering on the original wideband sampling data to obtain the bandpass-filtered data; among which, , , This indicates the number of points corresponding to the spectrum currently being analyzed using FFT; S6, Perform complex modulation refinement spectrum analysis on the bandpass filtered data to obtain the bandpass frequency range. Data frequency; S7, responding to the bandpass frequency range Exceed Complete full-band signal frequency measurement; responds to the bandpass frequency range. Not exceeding Return to S4 and proceed to the next frequency range processing flow.

2. The method for rapid frequency measurement of multi-frequency signals with continuous segmentation and secondary identification according to claim 1, characterized in that, In S1: The low-pass filtering process uses an FIR digital low-frequency filter, with the cutoff frequency set to... .

3. The method for rapid frequency measurement of multi-frequency signals with continuous segmentation and secondary identification according to claim 1, characterized in that, In S2, frequency resolution , The value can be adjusted according to the actual system.

4. The method for rapid frequency measurement of multi-frequency signals with continuous segmentation and secondary identification according to claim 1, characterized in that, In S5, the bandpass digital filter is designed using the equal ripple optimal approximation method, and the bandpass digital filter adopts the FIR digital filter.

5. The method for rapid frequency measurement of multi-frequency signals with continuous segmentation and secondary identification according to claim 1, characterized in that, In S6, ZoomFFT, a complex modulation refined spectral analysis method, is used to perform fine analysis on the bandpass filtered data. With a resampling interval of D, the frequency resolution becomes... Where D is the sampling ratio, the N-point time-domain sequence obtained after resampling is subjected to Fast Fourier Transform, which improves the resolution by a factor of D.

6. A rapid frequency measurement device for multi-frequency signals with continuous segmented secondary identification, characterized in that, Including processor and storage media; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1 to 5.

7. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images