CN104749435A - DFT harmonic detection method without accumulated error sliding window - Google Patents

Abstract

The invention discloses a DFT harmonic detection method without an accumulated error sliding window. The DFT harmonic detection method includes: collecting voltage signals of a power grid, and utilizing digital phase-locked loop technology to enable voltage of the power grid to be synchronous; performing DFT forward transformation on collected load current signals x(t) through a sliding window iteration method to respectively acquire x-axis and y-axis components; determining instruction signals of hth harmonic waves, and utilizing DFT reverse transformation to extract each harmonic instruction. By the DFT harmonic detection method, the problem of extraction of realtime harmonic instructions of an active power filter is solved, and each harmonic component can be detected accurately and quickly to guarantee realtime compensation effect of the active power filter.

Description

A kind of DFT harmonic detecting method without the sliding window of cumulative errors

Technical field

The invention belongs to digital signal processing technique field, more specifically say, relate to a kind of DFT harmonic detecting method without the sliding window of cumulative errors.

Background technology

Along with the development of electric system, nonlinear load is applied on a large scale, and they inject a large amount of harmonic wave to electrical network, power system voltage current waveform is distorted more and more serious, has had a strong impact on the quality of power supply and Electrical Safety.Therefore, harmonic detecting and suppression have great reality and economic implications.

At present, the detection method based on frequency domain and the large class of the detection method based on time domain two can be substantially divided into about Harmonic in Power System detection method.Based on the detection method of frequency domain based on Fourier transform, comprise Fast Fourier Transform (FFT) method (FFT), discrete Fourier transformation method (DFT) and iterative Fourier transform method (RDFT).Harmonic detecting method based on Fourier transform can realize harmonic wave gradation and detect, easy to use, but there is the inherent delay of one-period, and real-time is poor; Iterative Fourier transform method, dynamic responding speed aspect is then better than fourier transform method, but needs for it additionally increases 2N variable storage space for the detection of every order harmonic components.Based on time harmonic detection method, comprise based on time domain instantaneous reactive power theory (p-q) detection method, based on first-harmonic synchronous rotation transformation (d-q) detection method, based on harmonic synchronous rotational transform detection method.Harmonic detecting method based on time domain belongs to real-time detection method, common dynamic fast response time, but there is the contradiction that subharmonic detects and calculated amount is large.

Summary of the invention

The object of the invention is to overcome the deficiencies in the prior art, a kind of DFT harmonic detecting method without the sliding window of cumulative errors is provided, solve Active Power Filter-APF (APF) to a difficult problem for real-time harmonic instruction fetch, each harmonic component can also be detected accurately and rapidly, to ensure the real-Time Compensation effect of APF simultaneously.

For achieving the above object, a kind of DFT harmonic detecting method without the sliding window of cumulative errors of the present invention, is characterized in that, comprise the following steps:

(1), mains voltage signal is gathered, by Digital Phase-Locked Loop Technology, synchronised grids voltage-phase and frequency;

(2), gather load current signal x (t), then by sliding window alternative manner, DFT direct transform is carried out to load current signal x (t), obtain x-axis component A respectively _hx(k) and y-axis component A _hy(k);

$A_{\mathrm{hx}}\left(k\right)=A_{\mathrm{hx}}(k-1)+\frac{2}{N}\·\cos \left(\frac{2\mathrm{\π}\·h\·k}{N}\right)\·(x\left(k\right)-x(k-N))---\left(1\right)$

$A_{\mathrm{hy}}\left(k\right)=A_{\mathrm{hy}}(k-1)+\frac{2}{N}\·\sin \left(\frac{2\mathrm{\π}\·h\·k}{N}\right)\·(x\left(k\right)-x(k-N))---\left(2\right)$

Wherein, x (n) is the sampled signal of signal x (t) containing harmonic wave, and n is sampled point number, and N is sampled point number in the primitive period;

(3) the signal x of h subharmonic command signal when kth is clapped, is determined _h(n)

$x_h\left(n\right)=A_{\mathrm{hx}}\left(k\right)\·\cos \left(\frac{2\mathrm{\π}\·h\·n}{N}\right)+A_{\mathrm{hy}}\left(k\right)\·\sin \left(\frac{2\mathrm{\π}\·h\·n}{N}\right)---\left(3\right)$

(4) signal x when, the described kth of step (3) being clapped _hn () carries out inverse transformation

Further, in described step (2), sliding window alternative manner to the method that load current signal x (t) carries out DFT direct transform is:

(2.1) in load current, arrange three sizes is the array of N, i.e. x [N], sin_table [N], cos_table [N], stores the amplitude x (n) of iterative window internal burden current signal sampled value, the sine value of corresponding phase and cosine value respectively;

(2.2), A is calculated according to formula (1) _hx(k)

(2.2.1), calculate and skid off the data of window, be i.e. most legacy data

Most legacy data=x [k] × cos_table [k], then with this most legacy data divided by N obtain the cycle average after most old value;

(2.2.2), the data slipping into window are calculated, i.e. latest data

Upgrade x [k], cos_table [k] by the sampled value of current flow signal, obtain x [k] ', cos_table [k] '

Latest data=x [k] × cos_table [k] ', then with this latest data divided by N obtain the cycle average after last look;

(2.2.3), calculate kth and clap x-axis component A _hx(k), kth bat value=kth-1 bat value-most old value+last look;

(2.3), according to formula (2), calculate kth by the method for step (2.2.1) ~ (2.2.3) and clap y-axis component A _hyk (), wherein, replaces cos_table [k] when concrete calculating with sin_table [k].

Goal of the invention of the present invention is achieved in that

The invention discloses a kind of DFT harmonic detecting method without the sliding window of cumulative errors, first gather mains voltage signal, recycling Digital Phase-Locked Loop Technology makes line voltage synchronous; By sliding window process of iteration, DFT direct transform is carried out to load current signal x (t) gathered, obtain x-axis and y-axis component respectively; Finally determine the command signal of h subharmonic, recycling DFT inverse transformation extracts each harmonic instruction.The invention solves a difficult problem for Active Power Filter-APF real-time harmonic instruction fetch, each harmonic component can also be detected accurately and rapidly, to ensure the real-Time Compensation effect of active power filter simultaneously.

Meanwhile, a kind of DFT harmonic detecting method without the sliding window of cumulative errors of the present invention also has following beneficial effect:

(1), for the error accumulation that sliding window process occurs, the method that the present invention proposes is: the amplitude x (n) of sampled value and sine value, the cosine value of respective phase that store current signal in a window, and utilize the data source window stored, avoid error accumulation;

(2), the present invention conveniently can realize the extract real-time of subharmonic instruction;

(3), adopt look-up table to calculate sine and cosine value, save controller operation time;

(4), with C language programming, dsp chip realizes, implements simple and easy.

Accompanying drawing explanation

Fig. 1 is the testing process block diagram of h subharmonic;

Fig. 2 is sliding-window iterative algorithm schematic diagram;

Fig. 3 is the 3 subharmonic test experience figure not adopting this algorithm;

Fig. 4 is 3 subharmonic test experience figure after adopting this algorithm.

Embodiment

Below in conjunction with accompanying drawing, the specific embodiment of the present invention is described, so that those skilled in the art understands the present invention better.Requiring particular attention is that, in the following description, when perhaps the detailed description of known function and design can desalinate main contents of the present invention, these are described in and will be left in the basket here.

Embodiment

Fig. 1 is the testing process block diagram of h subharmonic.

For ease of describing, the present embodiment, for single-phase load currents, describes the concrete grammar of 3,5,7, the 11 subharmonic instructions of extracting single-phase load currents x (t).

In the present embodiment, as shown in Figure 1, extract the instruction of h subharmonic, comprise the following steps:

(1), create sine table, 0 ~ 2 π is divided into 200 (supposing that in the primitive period, sampling number is 200) equal portions, calculate the sine value of respective angles, and set up the sine table of 200 points accordingly;

(2), sampling mains voltage signal, by Digital Phase-Locked Loop Technology, synchronised grids voltage-phase θ;

(3), sampling single-phase load currents signal x (t), then by sliding window alternative manner, DFT direct transform is carried out to single-phase load currents signal x (t), obtains x-axis component A respectively _hx(k) and y-axis component A _hy(k);

$A_{\mathrm{hx}}\left(k\right)=A_{\mathrm{hx}}(k-1)+\frac{2}{N}\·\cos \left(\frac{2\mathrm{\π}\·h\·k}{N}\right)\·(x\left(k\right)-x(k-N))---\left(1\right)$

$A_{\mathrm{hy}}\left(k\right)=A_{\mathrm{hy}}(k-1)+\frac{2}{N}\·\sin \left(\frac{2\mathrm{\π}\·h\·k}{N}\right)\·(x\left(k\right)-x(k-N))---\left(2\right)$

Wherein, x (n) is the sampled signal of signal x (t) containing harmonic wave, and n is sampled point number, and N is sampled point number in the primitive period, i.e. N=200, k ∈ [1, n];

In the present embodiment, as shown in Figure 2, sliding window process of iteration to the concrete steps that single-phase load currents signal x (t) carries out DFT direct transform is:

(3.1) in load current x (t), arrange three length is the array of 200, i.e. x [200], sin_table [200], cos_table [200], stores the amplitude x (n) of iterative window internal burden current signal sampled value, the sine value of corresponding phase and cosine value respectively;

(3.2), A is calculated according to formula (1) _hx(k)

(3.2.1), calculate and skid off the data of window, be i.e. most legacy data

Most legacy data=x [k] × cos_table [k], then with this most legacy data divided by 200 obtain the cycle average after most old value;

(3.2.2), the data slipping into window are calculated, i.e. latest data

X [k], cos_table [k] is upgraded by the sampled value of current flow signal, obtain x [k] ', cos_table [k] ', latest data=x [k] × cos_table [k] ', then with this latest data divided by 200 obtain the cycle average after last look;

(3.2.3), calculate kth and clap x-axis component A _hx(k), kth bat value=kth-1 bat value-most old value+last look;

(3.3), according to formula (2), calculate kth by the method for step (3.2.1) ~ (3.2.3) and clap y-axis component A _hy(k), wherein, replace cos_table [k] when concrete calculating with sin_table [k], its concrete calculation process does not repeat them here;

(4) the signal x of h subharmonic command signal when kth is clapped, is determined _h(n)

$x_h\left(n\right)=A_{\mathrm{hx}}\left(k\right)\·\cos \left(\frac{2\mathrm{\π}\·h\·n}{200}\right)+A_{\mathrm{hy}}\left(k\right)\·\sin \left(\frac{2\mathrm{\π}\·h\·n}{200}\right)---\left(3\right)$

In the present embodiment, h value is 3,5,7,11;

(5) signal x when, the described kth of step (4) being clapped _hn () carries out inverse transformation

$x\left(n\right)=\underset{h=\mathrm{3,5,7,11}}{\mathrm{\Σ}}{A}_{\mathrm{hx}}\left(k\right)\·\cos \left(\frac{2\mathrm{\π}\·h\·n}{200}\right)+\underset{h=\mathrm{3,5,7,11}}{\mathrm{\Σ}}{A}_{\mathrm{hy}}\left(k\right)\·\sin \left(\frac{2\mathrm{\π}\·h\·n}{200}\right)---\left(4\right)$

In the present embodiment, gradation is extracted 3,5,7,11 subharmonic instructions of single-phase load currents signal x (t), then total harmonic wave instruction is x (n)=x ₃(n)+x ₅(n)+x ₇(n)+x ₁₁(n).

The harmonic wave that the present invention is applicable to threephase load electric current too extracts, and in threephase load electric current, extracts respectively according to the method described above, do not repeat them here a, b, c three-phase.

Fig. 3 is the 3 subharmonic test experience figure not adopting this algorithm;

Fig. 4 is 3 subharmonic test experience figure after adopting this algorithm.

Fig. 3 and Fig. 4 represents respectively and does not adopt and adopt patent load current 3 subharmonic test experience oscillogram of the present invention.In the present embodiment, input signal is peak-to-peak value is 2V, frequency is the square wave of 50Hz, when to identical load current, when not adopting the inventive method, because error can be accumulated in time, cause harmonic wave forms to be dispersed, can not accurately detect 3 subharmonic, as shown in Figure 3, the amplitude of harmonic wave is bigger than normal, is just embodying harmonic wave forms and is dispersing.As shown in Figure 4, after adopting method of the present invention, do not have error accumulation, harmonic wave forms is stable, and therefore the present invention can extract harmonic wave instruction quickly and accurately.

Although be described the illustrative embodiment of the present invention above; so that those skilled in the art understand the present invention; but should be clear; the invention is not restricted to the scope of embodiment; to those skilled in the art; as long as various change to limit and in the spirit and scope of the present invention determined, these changes are apparent, and all innovation and creation utilizing the present invention to conceive are all at the row of protection in appended claim.

Claims (2)

1., without a DFT harmonic detecting method for the sliding window of cumulative errors, it is characterized in that, comprise the following steps:

(1), mains voltage signal is gathered, by Digital Phase-Locked Loop Technology, synchronised grids voltage-phase and frequency;

(2), gather load current signal x (t), then by sliding window alternative manner, DFT direct transform is carried out to load current signal x (t), obtain x-axis component A respectively _hx(k) and y-axis component A _hy(k);

$A_{\mathrm{hx}}\left(k\right)=A_{\mathrm{hx}}(k-1)+\frac{2}{N}\·\cos \left(\frac{2\mathrm{\π}\·h\·k}{N}\right)\·(x\left(k\right)-x(k-N))---\left(1\right)$

$A_{\mathrm{hy}}\left(k\right)=A_{\mathrm{hy}}(k-1)+\frac{2}{N}\·\sin \left(\frac{2\mathrm{\π}\·h\·k}{N}\right)\·(x\left(k\right)-x(k-N))---\left(2\right)$

Wherein, x (n) is the sampled signal of signal x (t) containing harmonic wave, and n is for adopting sampled point number, and N is sampled point number in the primitive period, k ∈ [1, n];

(3) the signal x of h subharmonic command signal when kth is clapped, is determined _h(n)

$x_h\left(n\right)=A_{\mathrm{hx}}\left(k\right)\·\cos \left(\frac{2\mathrm{\π}\·h\·n}{N}\right)+A_{\mathrm{hy}}\left(k\right)\·\sin \left(\frac{2\mathrm{\π}\·h\·n}{N}\right)---\left(3\right)$

(4) signal x when, the described kth of step (3) being clapped _hn () carries out inverse transformation

$x\left(n\right)=\underset{h=1}{\overset{\∞}{\mathrm{\Σ}}}{A}_{\mathrm{hk}}\left(k\right)\·\cos \left(\frac{2\mathrm{\π}\·h\·n}{N}\right)+\underset{h=1}{\overset{\∞}{\mathrm{\Σ}}}{A}_{\mathrm{hy}}\left(k\right)\·\sin \left(\frac{2\mathrm{\π}\·h\·n}{N}\right).---\left(4\right)$

2. a kind of DFT harmonic detecting method without the sliding window of cumulative errors according to claim 1, is characterized in that, in described step (2), sliding window alternative manner to the method that load current signal x (t) carries out DFT direct transform is:

(2.1), in every phase current of load current, arranging three sizes is the array of N, i.e. x [N], sin_table [N], cos_table [N], stores the amplitude x (n) of iterative window internal burden current signal sampled value, the sine value of corresponding phase and cosine value respectively;

(2.2), A is calculated according to formula (1) _hx(k)

(2.2.1), calculate and skid off the data of window, be i.e. most legacy data

Most legacy data=x [k] × cos_table [k], then with this most legacy data divided by N obtain the cycle average after most old value;

(2.2.2), the data slipping into window are calculated, i.e. latest data

Upgrade x [k], cos_table [k] by the sampled value of current flow signal, obtain x [k] ', cos_table [k] '

Latest data=x [k] × cos_table [k] ', then with this latest data divided by N obtain the cycle average after last look;

(2.2.3), calculate kth and clap x-axis component A _hx(k), kth bat value=kth bat value-most old value+last look;

(2.3), according to formula (2), calculate kth by the method for step (2.2.1) ~ (2.2.3) and clap y-axis component A _hyk (), wherein, replaces cos_table [k] when concrete calculating with sin_table [k].