CN104714075A - Method for extracting power grid voltage flicker envelope parameters - Google Patents

Abstract

The invention discloses a method for extracting power grid voltage flicker envelope parameters. According to the method, power grid voltage flicker waveforms are sampled at a constant sampling frequency, a flicker envelope amplitude modulated wave component is obtained through energy calculation of an improved Teager energy operator, a cosine window is added to conduct Chirp-Z conversion, the flicker envelope amplitude modulated wave frequency and a correction factor of an amplitude modulated wave amplitude value are extracted by improving the Chirp-Z conversion, and the flicker envelope amplitude modulated wave amplitude value is corrected through the correction factor and accurately obtained; through the combination of the improvement on the energy operator and the improvement on the Chirp-Z conversion, the energy operator is improved, the flicker envelope demodulation calculation process by the energy operator is improved, the noise resistance is improved; by improving the Chirp-Z conversion, the flicker frequency analysis range is flexible and adjustable, and frequency spectrum leakage generated under the condition of asynchronous sampling and errors caused by a picket fence effect are effectively avoided; the correction factor of the amplitude modulated wave amplitude value reduces the errors caused by the energy operator, and the power grid voltage flicker envelope parameters can be accurately measured in real time.

Description

A kind of electric network voltage flicker envelope parameters extracting method

Technical field

The invention belongs to signal processing technology field, particularly a kind of electric network voltage flicker envelope parameters extracting method.

Background technology

Electric load, especially impact load sharply increase, and constitute serious pollution to electric system, cause line voltage unstable, produce voltage fluctuation and flicker, add the labile factor of electrical network, cause have a strong impact on to commercial production and social life.Voltage flicker is the important parameter of the quality of power supply, is the major reason caused for, consumer fault and inefficacy.Carry out Measurement accuracy to for the voltage flicker in, using electricity system, can be research flickering root, the impact etc. of suppression and elimination voltage fluctuation and flicker provides scientific basis.

The accurate estimation of voltage flicker envelope parameters is the most critical factor calculating flickering value and assessment flickering serious interference, many methods have been proposed for the acquisition of voltage flicker envelope parameters, and wherein widely used is discrete Fourier transformation (Fast Fourier Transform (FFT) FFT can be used to realize it calculate fast).But signal is under non-synchronous sampling, there is intrinsic spectral leakage and fence effect in FFT, and measurement result is by sample frequency and the restriction in sampling time.Be necessary to consider to increase the sampling time to improve frequency resolution, but be unaccommodated for unstable signal.Directly solution conditioning technology (IDDM) is widely used envelope demodulation method in FFT, but it needs square root to calculate, and is vulnerable to the impact of humorous wave interference and mains frequency skew.In order to overcome the restriction of FFT method, Kalman filtering (KF), least absolute value estimate that (LAV), continuous wavelet transform (CWT) and S-transformation have been applied to voltage flicker parameter estimation.For the application of Kalman filtering, be limited to the accurate estimation of its model parameter, and there is the large problem of calculated amount in it, is difficult in embedded realization.Least absolute value method of estimation hypothesis voltage flicker frequency is known, under flicker frequency variation condition, is difficult to obtain measurement result accurately, and also there is the large problem of calculated amount.Continuous wavelet transform is difficult to select suitable wavelet basis function, obtains parameter model difficulty accurately, there is the problem that calculated amount is large equally.In recent years, S-transformation has time frequency analysis ability because of it, be applied to flickering and analyze, but the method not only exists the large problem of calculated amount but also is limited to the impact of frequency resolution, is difficult to accurately obtain voltage flicker parameter.

Teager energy operator because of its simply, fast and accurately advantage be widely used flicker envelope and followed the trail of, but there is high frequency error in it, and easily affected by noise.Under the condition not changing sample frequency and sampling time, Chirp-Z conversion improves frequency resolution and reduces measuring error, has been applied in flickering analysis.But still there is spectral leakage and fence effect that square wave truncated signal under non-synchronous sampling causes in the method, is difficult to accurately extract flickering parameter, particularly contiguous modulated wave component.

Summary of the invention

For overcoming the deficiency of prior art, method, the invention provides a kind of voltage flicker envelope parameters extracting method, the method can effectively reduce the error that spectral leakage and fence effect produce under the error of Teager energy operator and non-synchronous sampling condition, overcome mains frequency fluctuation, harmonic wave and m-Acetyl chlorophosphonazo and noise to the impact of measurement result, realize fast in real time, the extraction of the voltage fluctuation envelope parameters of pin-point accuracy.

A kind of electric network voltage flicker envelope parameters extracting method, comprises following step:

Step 1: tested voltage flicker signal to be sampled and analog to digital conversion obtains discrete signal u (n), modified Teager energy operator is utilized to carry out energy calculation to discrete signal, obtain the discrete series ψ [u (n)] containing flicker envelope modulated wave component, voltage flicker signal sampling frequency is f _swith sampling number N;

The discrete form of described modified Teager energy operator is:

ψ[x(n)]＝x ²(n)-x(n-k)x(n+k) (1)

Wherein, n is discrete sampling times, n-k and n+k represents a front and back k sampling point value of this sampled point moment n respectively, and the span of k is 1 ~ 8;

[flickering refers to the intuitive visual impression that human eye is abnormal to the illumination caused by the voltage fluctuation of certain frequency and produce, and is caused by the amplitude fluctuations of line voltage changes.Usually flickering is regarded as with the power-frequency voltage in electrical network for carrier wave, using voltage fluctuation component as modulated wave, modulate the result of this carrier voltage root mean square or peak value.

Teager energy operator is a kind of nonlinear operator that Teager proposes when pronunciation modeling, amplitude and the angular frequency change of quick tracking signal is only got final product with three sampling points of signal, to signal frequency and amplitude trace ability strong, but all changes of signal are all reflected in energy operator, therefore it is to noise and voltage jump sensitivity.

In the technical program, selected modified Teager energy operator is not adopt three closely adjacent points to calculate, but the number of winning the confidence is separated by, k calculates.】

Step 2: utilize cosine time domain window w (n) to be weighted process to discrete series ψ [u (n)], obtain discrete series s (n) after weighting;

Step 3: the analysis of modified Chirp-Z conversion spectrum is done to discrete series s (n) after weighting and obtains discrete spectrum X _cZT(k);

Wherein, the analysis of described modified Chirp-Z conversion spectrum refers to that discrete series s (n) to length is N carries out analysis of spectrum, and its discrete spectrum expression formula is:

$X_{\mathrm{CZT}}\left(k\right)=\frac{m_i}{2j}{e}^{j{\mathrm{\θ}}_i}W[2\mathrm{\π}(k-k_i)/N_Z]---\left(2\right)$

[second harmonic component is ignored in this Chirp-Z conversion and other flicker envelope modulated wave components disturb the leakage of i-th flicker envelope modulated wave component; ]

In formula, θ _iit is the initial phase angle of i-th flicker envelope modulated wave; Nz is modified Chirp-Z transform data length, N _z=f _sm/ (f _h-f _l), [f _l, f _h] be the frequency range analysis scope that Chirp-Z converts, M is that the segmentation that modified Chirp-Z converts is counted, and W () is the discrete spectrum function of cosine time domain window;

Step 4: to the discrete spectrum X obtained _cZTk () obtains flicker envelope modulated wave frequency f by the method for approximation by polynomi-als _icorrection formula, thus obtain the correction factor K of flicker envelope modulated wave amplitude _i;

Described flicker envelope modulated wave frequency f _ifollowing formula of revising is adopted to calculate acquisition:

f _i＝k _iΔf _z+f _l＝(η+k _i1+0.5)Δf _z+f _l(3)

[due to the fence effect that non-synchronous sampling causes, the crest frequency k of i-th flicker envelope modulated wave in flicker envelope _iΔ f _z+ f _lbe difficult to be on discrete spectral line frequency, i.e. k _igeneral is not integer; ]

Wherein, peak point k _ineighbouring amplitude maximum and secondary maximum spectral line k respectively _i1and k _i2, k _i1≤ k _i≤ k _i2=k _i1+ 1, the amplitude that maximum and secondary maximum spectral line is corresponding is respectively y _i1=| X _cZT(k _i1) | and y _i2=| X _cZT(k _i2) |, k _ifor flicker envelope modulated wave frequency f _ithe position of corresponding frequency spectrum, k _i=(f _i-f _l)/Δ f _z, Δ f _zfor frequency resolution, Δ f _z=(f _h-f _l)/M; η is flicker envelope modulated wave frequency spectrum position adjustments parameter, η=k _i-k _i1-0.5, η span is [-0.5,0.5], and ξ is that setup parameter is asked for, ξ=(y in flicker envelope modulated wave frequency spectrum position _i1-y _i2)/(y _i1+ y _i2);

And η=h ^-1(ξ)=H (ξ), wherein,

$\mathrm{\ξ}=h\left(\mathrm{\η}\right)=\frac{\left|W\left(\frac{2\mathrm{\π}(-\mathrm{\η}+0.5)}{N_z}\right)\right|-\left|W\left(\frac{2\mathrm{\π}(-\mathrm{\η}-0.5)}{N_z}\right)\right|}{\left|W\left(\frac{2\mathrm{\π}(-\mathrm{\η}+0.5)}{N_z}\right)\right|+\left|W\left(\frac{2\mathrm{\π}(-\mathrm{\η}-0.5)}{N_z}\right)\right|}---\left(4\right)$

η is got a class value in [-0.5,0.5], is obtained the value of corresponding ξ by ξ=h (η), call polyfit (η, ξ, ρ) function and carry out anti-fitting, ρ is the polynomial exponent number of approach, obtains the coefficient of polynomial expression H (ξ);

The correction factor K of described flicker envelope modulated wave amplitude _i:

$K_i=2{\sin }^2\left(k{\mathrm{\Ω}}_0\right)+4{\sin }^2\left(\frac{k{\mathrm{\Ω}}_i}{2}\right)[\frac{1}{2}-{\sin }^2\left(k{\mathrm{\Ω}}_0\right)]---\left(5\right)$

In formula, Ω ₀=2 π f ₀/ f _s, Ω _i=2 π f _i/ f _s, f ₀for fundamental frequency;

Step 5: utilize described correction factor K _irevise the amplitude m obtaining flicker envelope modulated wave component accurately _i:

$m_i=(y_{i1}+y_{i2}){(N_Z\·U_0^2\·K_i)}^{-1}g\left(\mathrm{\η}\right)---\left(6\right)$

Wherein, g (η) represents the polynomial expression relevant with flicker envelope modulated wave frequency spectrum position adjustments parameter η;

η is got a class value, by formula in [-0.5,0.5] simplified style m _i=N _z ^-1(y _i1+ y _i2) g (η) draws corresponding g (η) value, calls the coefficient that polyfit (η, g (η), ρ) function obtains polynomial expression g (η).

Preferably, fixed sampling frequency f is adopted _sspan at 200Hz ~ 64KHz, within the scope of mains frequency 45 ~ 55Hz, f _ssynchronously do not change with mains frequency.

Preferably, described discrete series ψ [u (the n)] computing formula containing flicker envelope modulated wave component is as follows:

Wherein, U ₀, θ ₀be respectively amplitude and the initial phase angle of electrical network fundamental voltage; m _i, θ _i`be respectively coefficient of variation and the initial phase angle of i-th flicker envelope modulated wave; H is the item number of voltage flicker envelope modulated wave; f _ifor the frequency of flicker envelope modulated wave component; ψ ₁n () is DC component, ψ ₂n () is flicker envelope modulated wave component, ψ ₃n () is second harmonic component, ψ ₄n () is the second harmonic component of flicker envelope modulated wave which amplitude modulation.

[the simplification derivation of discrete series ψ [u (n)] is as follows:

The computing of improvement energy operator is carried out to discrete voltage flicker signal u (n), can obtain

In formula, D (n) and E (n) is respectively extraction factor and error factor.

For discrete form, for simplifying voltage flicker energy operator formula, launching the ψ [1+v (n)] in D (n) and E (n), having

$\begin{array}{c}D\left(n\right)={[1+v\left(n\right)]}^2{\sin }^2\left(k{\mathrm{\Ω}}_0\right)\\ =\left[\begin{array}{c}1+\underset{i=1}{\overset{h}{\mathrm{\Σ}}}{m_i}^2{\cos }^2({\mathrm{\Ω}}_{i}n+{\mathrm{\θ}}_i)+2\underset{i=1}{\overset{h}{\mathrm{\Σ}}}{m}_i\cos ({\mathrm{\Ω}}_{i}n+{\mathrm{\θ}}_i)\\ +2\underset{i,j=1,i\≠j}{\overset{h}{\mathrm{\Σ}}}{m}_i{m}_j\cos ({\mathrm{\Ω}}_{i}n+{\mathrm{\θ}}_i)\cos ({\mathrm{\Ω}}_{j}n+{\mathrm{\θ}}_j)\end{array}\right]\·{\sin }^2\left(k{\mathrm{\Ω}}_0\right)\end{array}---\left(9\right)$

$\begin{array}{c}\mathrm{\ψ}[1+v\left(n\right)]={[1+v\left(n\right)]}^2-[1+v(n+k)][1+v(n-k)]\\ ={[1+\underset{i=1}{\overset{h}{\mathrm{\Σ}}}{m}_i\cos ({\mathrm{\Ω}}_{i}n+{\mathrm{\θ}}_i)]}^2-\{1+\underset{i=1}{\overset{h}{\mathrm{\Σ}}}{m}_i\cos [{\mathrm{\Ω}}_i(n+k)+{\mathrm{\θ}}_i\left]\right\}.\\ \{1+\underset{j=1}{\overset{h}{\mathrm{\Σ}}}{m}_j\cos [{\mathrm{\Ω}}_j(n-k)+{\mathrm{\θ}}_i\left]\right\}\end{array}$

$\begin{array}{c}=\underset{i=1}{\overset{h}{\mathrm{\Σ}}}{m_i}^2{\sin }^2\left(k{\mathrm{\Ω}}_i\right)+4\underset{i=1}{\overset{h}{\mathrm{\Σ}}}{m}_i{\sin }^2(k{\mathrm{\Ω}}_i/2)\cos ({\mathrm{\Ω}}_{i}n+{\mathrm{\θ}}_i)\\ +\underset{i=j=1,i\≠j}{\overset{h}{\mathrm{\Σ}}}{m}_i{m}_j\left\{\begin{array}{c}\cos ({\mathrm{\Ω}}_{i}n+{\mathrm{\θ}}_i)\cos ({\mathrm{\Ω}}_{j}n+{\text{\θ}}_j)\·[1-\cos \left(k{\text{\Ω}}_i\right)\cos \left(k{\mathrm{\Ω}}_i\right)]\\ +\sin \left({\mathrm{\Ω}}_{i}n{+\mathrm{\θ}}_i\right)\sin ({\mathrm{\Ω}}_{j}n+{\mathrm{\θ}}_j)\·\sin \left(k{\mathrm{\Ω}}_i\right)\sin (k{\mathrm{\Ω}}_j\end{array}\right\}\end{array}---\left(10\right)$

Due to the amplitude U of flickering modulated wave signal _iusually electrical network fundamental voltage amplitude U is less than ₀10%, i.e. flicker envelope modulated wave Coefficient m _ibe less than 10%, flicker envelope modulated wave frequency f _ibe 0.05 ~ 35Hz, electrical network fundamental voltage frequency f ₀fluctuate near 50Hz, the item number h of flicker envelope modulated wave is usually less, has

$D\left(n\right)\≈[1+2\underset{i=1}{\overset{h}{\mathrm{\Σ}}}{m}_i\cos ({\mathrm{\Ω}}_{i}n+{\mathrm{\θ}}_i)]\·{\sin }^2\left(k{\mathrm{\Ω}}_0\right)---\left(11\right)$

$\mathrm{\ψ}[1+v\left(n\right)]\≈\underset{i=1}{\overset{h}{\mathrm{\Σ}}}{m_i}^2{\sin }^2\left(k{\mathrm{\Ω}}_i\right)+4\underset{i=1}{\overset{h}{\mathrm{\Σ}}}{m}_i{\sin }^2(\text{k}{\mathrm{\Ω}}_i/2)\cos ({\mathrm{\Ω}}_{i}n+{\mathrm{\θ}}_i)---\left(12\right)$

Formula (11), (12) are substituted into formula (8), obtains formula (7).】

Preferably, described cosine time domain window is:

$w\left(n\right)=\underset{l=0}{\overset{L-1}{\mathrm{\Σ}}}{(-1)}^l{b}_l\cos (2\mathrm{\πn}\·l/N)---\left(13\right)$

In formula: L is the item number of window function, and l represents l item; N=1,2,3 ... .N; b _lmeet constraint condition: $\underset{l=0}{\overset{L-1}{\mathrm{\Σ}}}{(-1)}^l{b}_l=0.$

Although [traditional Chirp-Z conversion improves frequency resolution, the rectangular window that signal cutout adopts, and its main lobe width is wide, sidelobe level is high, side lobe attenuation speed is slow, increases contiguous leakage, reduces measuring accuracy.

The remaining window function weighting of excellent main lobe and side lobe attenuation speed characteristic can effectively reduce the frequency caused because of non-synchronous sampling reveal and fence effect on the impact of measurement result, raising measuring accuracy.】

Preferably, described cosine time domain window 51 rank Rife-Vincent windows.

Beneficial effect

Compared with prior art, the present invention combines by improving energy operator and improving Chirp-Z conversion, improves the computation process that energy operator simplifies energy operator envelope demodulation, improves noise immunity; Improving Chirp-Z conversion, to achieve flicker frequency analyst coverage adjustable flexibly, improves frequency resolution, effectively overcomes the error that spectral leakage under non-synchronous sampling and fence effect produce; The flicker envelope modulated wave amplitude rectification factor proposed decreases the high frequency error that energy operator brings, and improves measuring accuracy; Overcome mains frequency variation, harmonic wave and m-Acetyl chlorophosphonazo and noise to the impact of measurement result simultaneously, algorithm design is flexible, calculating is simple, achieve the real-time Measurement accuracy of electric network voltage flicker envelope parameters, for research flickering impact, flicker source location, flicker control and elimination provide effective foundation.

Accompanying drawing explanation

Fig. 1 is the measuring table schematic diagram of a kind of electric network voltage flicker envelope parameters extracting method that the embodiment of the present invention provides;

Fig. 2 is 51 rank Rife-Vincent window normalization log spectrum figure;

Fig. 3 is the process flow diagram of a kind of electric network voltage flicker envelope parameters extracting method that the embodiment of the present invention provides;

Fig. 4 is the measuring process figure of a kind of electric network voltage flicker envelope parameters extracting method that the embodiment of the present invention provides, wherein, a () is 2s sampled voltage waveform schematic diagram, (b) is voltage flicker envelope waveform schematic diagram, and (c) is CZT spectrum diagram.

Embodiment

The test platform of a kind of electric network voltage flicker envelope parameters extracting method that the embodiment of the present invention provides as shown in Figure 1, signal generator uses the Aglilent 33500B of Agilent company of the U.S. to produce flickering signal u (n), association ThinkpadE40 notebook computer connection signal generator is used by PC, the parameter that signal occurs is controlled, data acquisition uses the NI data collecting card ELVIS II of National Instruments, the data obtained is obtained by the capture program of the LabVIEW software of National Instruments, and the signal of generation is monitored in real time by the oscillograph Agilent DSO1102B of Agilent company of the U.S..Finally, the data importing DSP program collected by LabVIEW also carries out simulation run by the CCS software of Texas Instruments, and wherein dsp processor uses the TMS320C6745 of Texas Instruments.

Major parameter is selected as follows:

(1) sampling rate: f _s=400Hz;

(2) electrical network fundamental frequency: f ₀=50Hz;

(3) electrical network fundamental voltage amplitude: U ₀=4V;

(4) count in the interval of improving energy operator: k=2;

(5) CZT operational data length: N=2000;

(6) segmentation of CZT operation frequency is counted: M=1000;

(7) CZT operation frequency analyst coverage is at 0 ~ 36Hz.

In the present embodiment, the expression formula of 51 rank Rife-Vincent windows is as follows:

w _RV(n)＝1-1.6cos(2πn/N)+0.8cos(4πn/N) (14)

-0.22857cos(6πn/N)+0.02857cos(8πn/N)

In formula, N is the data length of CZT computing, is also length (n=1,2 of window function, ..., N-1), as shown in Figure 2, the progressive rate of decay of secondary lobe is 30dB/oct, side lobe peak level reaches-74.5dB, have desirable sidelobe performance, therefore, the present embodiment selects 51 rank Rife-Vincent windows to be weighted process.

Convert derivation by above-mentioned condition and improvement Chirp-Z can obtain, coefficient η, modulated wave frequency f _i, correction factor K _iwith modulated wave amplitude m _icorrection formula as follows:

η＝139.4246745191269ξ (15)

f _i＝k _iΔf＝(η+k _i1+0.5)Δf _z(16)

$K_i=2{\sin }^2(2\mathrm{\π}{f}_0/f_s)+4{\sin }^2(\mathrm{\π}{f}_i/f_s)\·[\frac{1}{2}-{\sin }^2(2\mathrm{\π}{f}_0/f_s)]---\left(17\right)$

m _i＝(y _i1+y _i2)(N _z·U ₀ ²·K _i) ^-1· (18)

(5.565523907094415+0.039813872470502η ²)

As shown in Figure 3, its concrete steps comprise the process flow diagram of a kind of electric network voltage flicker envelope parameters extracting method provided in the present embodiment:

(1) measured signal to be sampled and analog to digital conversion obtains discrete signal u (n), utilize modified Teager energy operator (improved Teager energy operator, ITEO) definition (1) carries out energy calculation to discrete signal, obtains such as formula the discrete series ψ [u (n)] containing modulated wave component shown in (7);

(2) 51 rank Rife-Vincent windows (five-term Rife-Vincent (I), FRVI) that utility is excellent are weighted process to discrete series;

(3) Discrete Linear Chirp-Z Transform (Chirp-Z) is done to the signal after weighting, obtain discrete spectrum X _cZT(k);

(4) due to the impact of spectral leakage under non-synchronous sampling condition, easily cause chaff component, particularly close on modulated wave component signal, set the chaff component in threshold value removal discrete spectrum for this reason;

(5) search obtains the amplitude spectral line y in spectrum sequence near extreme value _i1, y _i2;

(6) coefficient ξ, η is calculated respectively by formula (4) and (15);

(7) flicker envelope modulated wave frequency f is calculated respectively by formula (16), (17) _iwith correction factor K _i;

(8) last, flicker envelope modulated wave amplitude m _ithrough type (18) is by correction factor K _icorrect and obtain.

According to the measuring process figure that the method for the invention obtains, as shown in Figure 4, (a) is 2s sampled voltage waveform schematic diagram, and actual measured results is as shown in table 1.

Table 1

Shown in figure (b), improve the accurate demodulation of energy operator and obtain flicker envelope, the CZT result of spectrum analysis of the flicker envelope that demodulation obtains is as figure (c).From figure (c), the spectral leakage caused due to non-synchronous sampling and the impact of fence effect, the chaff component that amplitude is less is there is in frequency spectrum, therefore, the method of setting threshold value is adopted to remove these chaff component, the correction factor of flicker envelope modulated wave frequency and modulated wave amplitude is obtained again by the method for approximation by polynomi-als, correction factor corrects and obtains flicker envelope modulated wave amplitude accurately, and the measurement result of actual flicker envelope modulated wave frequency and amplitude demonstrates accuracy and the validity of the method for the invention.

Claims (5)

1. an electric network voltage flicker envelope parameters extracting method, is characterized in that, comprises following step:

Step 1: tested voltage flicker signal to be sampled and analog to digital conversion obtains discrete signal u (n), modified Teager energy operator is utilized to carry out energy calculation to discrete signal, obtain the discrete series ψ [u (n)] containing flicker envelope modulated wave component, voltage flicker signal sampling frequency is f _swith sampling number N;

The discrete form of described modified Teager energy operator is:

ψ[x(n)]＝x ²(n)-x(n-k)x(n+k)

Wherein, n is discrete sampling times, n-k and n+k represents a front and back k sampling point value of this sampled point moment n respectively, and the span of k is 1 ~ 8;

Step 2: utilize cosine time domain window w (n) to be weighted process to discrete series ψ [u (n)], obtain discrete series s (n) after weighting;

Step 3: the analysis of modified Chirp-Z conversion spectrum is done to discrete series s (n) after weighting and obtains discrete spectrum X _cZT(k);

Wherein, the analysis of described modified Chirp-Z conversion spectrum refers to that discrete series s (n) to length is N carries out analysis of spectrum, and its discrete spectrum expression formula is:

$X_{\mathrm{CZT}}\left(k\right)=\frac{m_i}{2j}{e}^{{\mathrm{j\θ}}_i}W[2\mathrm{\π}(k-k_i)/N_Z]$

In formula, θ _iit is the initial phase angle of i-th flicker envelope modulated wave; Nz is modified Chirp-Z transform data length, N _z=f _sm/ (f _h-f _l), [f _l, f _h] be the frequency range analysis scope that Chirp-Z converts, M is that the segmentation that modified Chirp-Z converts is counted, and W () is the discrete spectrum function of cosine time domain window;

Step 4: to the discrete spectrum X obtained _cZTk () obtains flicker envelope modulated wave frequency f by the method for approximation by polynomi-als _icorrection formula, thus obtain the correction factor K of flicker envelope modulated wave amplitude _i;

Described flicker envelope modulated wave frequency f _ifollowing formula of revising is adopted to calculate acquisition:

f _i＝k _iΔf _z+f _l＝(η+k _i1+0.5)Δf _z+f _l

Wherein, peak point k _ineighbouring amplitude maximum and secondary maximum spectral line k respectively _i1and k _i2, k _i1≤ k _i≤ k _i2=k _i1+ 1, the amplitude that maximum and secondary maximum spectral line is corresponding is respectively y _i1=| X _cZT(k _i1) | and y _i2=| X _cZT(k _i2) |, k _ifor flicker envelope modulated wave frequency f _ithe position of corresponding frequency spectrum, k _i=(f _i-f _l)/Δ f _z, Δ f _zfor frequency resolution, Δ f _z=(f _h-f _l)/M; η is flicker envelope modulated wave frequency spectrum position adjustments parameter, η=k _i-k _i1-0.5, η span is [-0.5,0.5], and ξ is that setup parameter is asked for, ξ=(y in flicker envelope modulated wave frequency spectrum position _i1-y _i2)/(y _i1+ y _i2);

And $\mathrm{\η}=h^{-1}\left(\mathrm{\ξ}\right)=H\left(\mathrm{\ξ}\right),\mathrm{\ξ}=h\left(\mathrm{\η}\right)=\frac{\left|W\left(\frac{2\mathrm{\π}(-\mathrm{\η}+0.5)}{N_2}\right)\right|-\left|W\left(\frac{2\mathrm{\π}(-\mathrm{\η}-0.5)}{N_z}\right)\right|}{\left|W\left(\frac{2\mathrm{\π}(-\mathrm{\η}+0.5)}{N_z}\right)\right|+\left|W\left(\frac{2\mathrm{\π}(-\mathrm{\η}-0.5)}{N_z}\right)\right|};$

η is got a class value in [-0.5,0.5], is obtained the value of corresponding ξ by ξ=h (η), call polyfit (η, ξ, ρ) function and carry out anti-fitting, ρ is the polynomial exponent number of approach, obtains the coefficient of polynomial expression H (ξ);

The correction factor K of described flicker envelope modulated wave amplitude _i:

$K_i=2{\sin }^2\left(k{\mathrm{\Ω}}_0\right)+4{\sin }^2\left(\frac{{\mathrm{k\Ω}}_i}{2}\right)[\frac{1}{2}-{\sin }^2\left({\mathrm{k\Ω}}_0\right)]$

In formula, Ω ₀=2 π f ₀/ f _s, Ω _i=2 π f _i/ f _s, f ₀for fundamental frequency;

Step 5: utilize described correction factor K _irevise the amplitude m obtaining flicker envelope modulated wave component accurately _i:

$m_i=(y_{i1}+y_{i2}){(N_Z\·U_0^2\·K_i)}^{-1}g\left(\mathrm{\η}\right)$

Wherein, g (η) represents the polynomial expression relevant with flicker envelope modulated wave frequency spectrum position adjustments parameter η;

η is got a class value, by formula in [-0.5,0.5] simplified style m _i=N _z ^-1(y _i1+ y _i2) g (η) draws corresponding g (η) value, calls the coefficient that polyfit (η, g (η), ρ) function obtains polynomial expression g (η).

2. a kind of electric network voltage flicker envelope parameters extracting method according to claim 1, is characterized in that, adopts fixed sampling frequency f _sspan at 200Hz ~ 64KHz, within the scope of mains frequency 45 ~ 55Hz, f _ssynchronously do not change with mains frequency.

3. a kind of electric network voltage flicker envelope parameters extracting method according to claim 1, is characterized in that, described discrete series ψ [u (the n)] computing formula containing flicker envelope modulated wave component is as follows:

Wherein, U ₀, θ ₀be respectively amplitude and the initial phase angle of electrical network fundamental voltage; m _i, θ _i`be respectively coefficient of variation and the initial phase angle of i-th flicker envelope modulated wave; H is the item number of flicker envelope modulated wave; f _ifor the frequency of flicker envelope modulated wave component; ψ ₁n () is DC component, ψ ₂n () is flicker envelope modulated wave component, ψ ₃n () is second harmonic component, ψ ₄n () is the second harmonic component of flicker envelope modulated wave which amplitude modulation.

4. a kind of electric network voltage flicker envelope parameters extracting method according to claim 1, is characterized in that, described cosine time domain window is:

$w\left(n\right)=\underset{l=0}{\overset{L-1}{\mathrm{\Σ}}}{(-1)}^l{b}_l\cos (2\mathrm{\πn}\·l/N)$

In formula: L is the item number of window function, and l represents l item; N=1,2,3 ... .N; b _lmeet constraint condition: $\underset{l=0}{\overset{L-1}{\mathrm{\Σ}}}{(-1)}^l{b}_l=0.$

5. a kind of electric network voltage flicker envelope parameters extracting method according to claim 4, is characterized in that, described cosine time domain window 51 rank Rife-Vincent windows.