CN102879639A - Real-time frequency measuring method in power system - Google Patents

Abstract

The invention discloses a real-time frequency measuring method in a power system. The method is used for tracking the frequency of a signal waveform to be measured in the power system in real time. The method comprises the following steps of: (1) establishing a rectangular coordinate system according to the waveform to be measured, wherein a horizontal axis is time, a longitudinal axis is a signal instantaneous value, and an intersection point of the longitudinal axis and the horizontal axis is a theoretical zero crossing point; (2) continuously sampling the waveform to be measured, and searching two continuous sampling points which fall on two sides of a random first theoretical zero crossing point, and two sampling points which are positioned behind the first theoretical zero crossing point and adjacent to the first theoretical zero crossing point and have the same change rate symbols; (3) linearly connecting the two sampling points respectively according to sampling values, and determining the intersection points of the sampling points and the horizontal axis as approximate zero crossing points; and (4) according to two determined approximate zero crossing points, measuring the frequency of the signal waveform to be measured. The frequency measuring method is simplicity in measurement, small in operation amount, high in real-time performance, high in accuracy and applicable to the occasions with requirements on high real-time performance and high measurement accuracy.

Description

A kind of method for real-time measurement of electric system medium frequency

Technical field

The invention belongs to the frequency measurement technical field in the electric system, specifically a kind of method for real-time measurement of electric system medium frequency.

Background technology

Frequency is one of the most basic electrical quantities, and the measurements such as the voltage in the electric system, electric current, power quality analysis all depend on the accurate measurement to fundamental frequency.Because the frequency of electric system reality fluctuates, therefore must follow the tracks of the frequency of fluctuation (namely measure in real time, or accurate synchro measure) just can carry out accurate electric current, voltage measurement and power quality analysis.

At present the method for survey frequency is a lot, as based on the periodic method of zero crossing, error minimize principle class method, based on the analytical method of signal observation model, DFT algorithm etc.Based on the periodic method of zero crossing be the measuring-signal waveform in succession the time width between zero crossing come survey frequency, the method physical concept is clear, calculated amount is little, but measuring accuracy is lower, and is subject to the impact of harmonic wave, random disturbance and aperiodic component.Error minimize principle class method is to realize the best-fit of sample data and model under the minimum variance meaning, the method fast convergence rate, to suppress the noise ability stronger, certain measuring accuracy is arranged, but when frequency shift (FS) 50Hz, there is the asynchronous sampling error in the frequency of measuring.Analytical method based on the signal observation model is to suppose that the effective information in the input signal meets a certain definite model, makes the input sample data be fit to greatest extent this model, and this algorithm has complicated mathematical operation, and amount of calculation is large.The DFT algorithm is first-harmonic and the harmonic separation of signal, thereby obtains the fundamental frequency of signal, and the precision and stability of this algorithm is better, but noise is larger on the measuring accuracy impact, and calculated amount is larger, and may the pull-in frequency aliasing to the periodic extension of signal.

CN102236048A discloses a kind of measuring method of electric system phasor frequency, it carries out low-pass filtering through low-pass filter first to the original interval discrete sampling value sequence of fixing time, with the interference of secondary and the above harmonic component of secondary in the elimination original input signal; Again to calculating respectively real part and imaginary part from the sampled value sequence, thereby the phase place of calculating calculates revised current high precision frequency values according to formula at last.The method is not subjected to the interference of power supply reference level fluctuation, and can not continue the impact of the calculated value of being subjected to over.But there is the defective that measuring accuracy is lower, calculated amount is excessive and requirement of real-time is not high in this measuring method.

Summary of the invention

The object of the invention is to propose a kind of method for real-time measurement of electric system medium frequency, by the zero crossing detection of similar triangle theory, in conjunction with filtering peace equalization process, realize the real-time measurement of high-precision frequency.

Realize that the concrete technical scheme that purpose of the present invention adopts is as follows:

A kind of method for real-time measurement of electric system medium frequency, the frequency for real-time follow-up measured signal waveform is characterized in that, comprising:

(1) take the time value of measured signal waveform as transverse axis, instantaneous value is the longitudinal axis, on this signal waveform to be measured, sets any two adjacent and theoretical zero crossing O that the rate of change symbol is identical ₁, O ₂

(2) two adjacent sampled points of supposition drop on respectively the both sides of theoretical zero crossing, obtain zero crossing O ₁Two sampled point u of both sides _I-1, u _i, and zero crossing O ₂Two sampled point u of both sides _J-1, u _jI, j are respectively the time series number of sampled point.

Wherein, two sampled point u _I-1, u _iBetween line and the intersection point of time shaft be B ₁, two sampled point u _J-1And u _jLine and the intersection point of time shaft be B ₂, this intersection points B ₁And B ₂Be approximate zero crossing.

(3) according to above-mentioned definite approximate zero crossing, can obtain the frequency f of signal waveform to be measured, namely

$f=\frac{1}{(j-i)\×T_s+t_{B_1{C}_1}-t_{B_2{C}_2}}$

In the formula, T _SBe sampling interval,

Be approximate zero crossing B ₁With sampled point u _iCorresponding time point C ₁Between the time interval,

Be approximate zero crossing B ₂With sampled point u _jCorresponding time point C ₂Between the time interval.

When sample frequency enough when high (being that sampling time interval is very little), line adjacent 2 on the curve is approximately straight line, and sampling interval is less, and curve more approaches straight line, and is more accurate by frequency or cycle that the method is determined.

In the method for the present invention, waveform to be measured must be processed through digital filtering, and the interference of filtering direct current and harmonic wave is to obtain accurate zero crossing.

Frequency measurement method of the present invention have measure simple, operand is little, real-time good, the enough highs of precision, is suitable for the requirement real-time good, the occasion that measuring accuracy is higher.

Description of drawings

Fig. 1 is the principle schematic that the method for embodiments of the invention is carried out frequency measurement.

Specific implementation method

Below in conjunction with the drawings and specific embodiments the present invention is described in detail.Need to prove, following embodiment only is exemplary, and not consisting of is limitation of the invention.

Take the sine voltage signal as example, specify the process of carrying out frequency measurement in the present embodiment, but the measuring-signal in the present embodiment is not defined as voltage signal, the various signals in the electric system such as electric current, electric field, magnetic field etc. are all applicable.

Such as Fig. 1, the method for real-time measurement of the frequency of the signal waveform in the present embodiment specifically comprises the steps:

(1) waveform voltage signal corresponding to be measured is set up corresponding rectangular coordinate system, and wherein the coordinate system transverse axis is the time, and the longitudinal axis is instantaneous voltage, and waveform is symmetrical about transverse axis.

Magnitude of voltage on the longitudinal axis with respect to transverse axis value on one side be on the occasion of, the another side value is negative value, voltage waveform and transverse axis intersection point are the voltage null value, this intersection point is also referred to as zero crossing.The time interval between any two adjacent and the zero crossings that the rate of change symbol is identical is the cycle of waveform, can obtain the frequency of waveform according to this waveform.In fact, because waveform is in continuous fluctuation, theoretical zero crossing is difficult to accurately obtain, and can only by obtaining as far as possible near the point of this theory zero crossing, carry out the measurement of wave period and frequency with this.

(2) voltage waveform is sampled, obtain a plurality of sampled points on the waveform, to each theoretical zero crossing, further find two adjacent sampled points of its both sides of apportion.

For any two adjacent and theoretical zero crossing O that the rate of change symbol is identical ₁, O ₂, suppose to lay respectively at zero crossing O ₁Two sampled points of both sides are u _I-1, u _i, lay respectively at zero crossing O ₂Two sampled points of both sides are u _J-1, u _j, wherein i, j are respectively the time series number of sampled point.

Such as Fig. 1, u _I-1And u _iLay respectively at zero crossing O ₁Left and right sides, its corresponding point on waveform are respectively D ₁And E ₁, u _J-1And u _jLay respectively at zero crossing O ₂Left and right sides, its corresponding point on waveform are respectively D ₂And E ₂

(3) determine the approximate zero crossing of corresponding each theoretical zero crossing according to the sampled point that finds.

Sampled point u _I-1And u _iTime point corresponding on time shaft is respectively A ₁And C ₁, sampled point u _J-1And u _jCorresponding time point is A respectively ₂And C ₂

Next, connect two sampled point u _I-1And u _iBetween line be D ₁E ₁, and two sampled point u _J-1And u _jBetween line be D ₂E ₂, simultaneously, tie-time point A ₁And C ₁Form line segment A ₁C ₁, tie-time point A ₂And C ₂Form line segment A ₂C ₂, and obtain respectively D ₁E ₁With A ₁C ₁Intersection points B ₁, D ₂E ₂With A ₂C ₂Intersection points B ₂

Wherein, line segment A ₁C ₁And A ₂C ₂Be positioned on the time shaft t, overlap with time shaft.This intersection points B ₁And B ₂The approximate zero crossing of definition signal waveform;

(4) according to above-mentioned definite approximate zero crossing B ₁And B ₂, can obtain the frequency f of signal waveform to be measured, namely

$f=\frac{1}{(j-i)\×T_s+t_{B_1{C}_1}-t_{B_2{C}_2}}$

In the formula, T _SBe sampling interval,

Be approximate zero crossing B ₁With sampled point u _iCorresponding time point C ₁Between the time interval,

Be approximate zero crossing B ₂With sampled point u _iCorresponding time point C ₂Between the time interval.

In this step (4), frequency formula obtains by the similar triangles method.

Be respectively u in the sinusoidal wave zero passage O1 of place, O2 left and right sides sampled value _I-1, u _iAnd u _J-1, u _j, available straight line D ₁B ₁E ₁Approximate D therebetween ₁O ₁E ₁Sinusoidal segments, with D ₂B ₂E ₂Approximate D therebetween ₂O ₂E ₂Sinusoidal segments, with B ₁, B ₂Replace O ₁, O ₂, the error that causes is less.When sample frequency enough when high (being that sampling time interval is enough little), line adjacent 2 on the curve is approximately straight line, and sampling interval is less, and straight line more approaches curve, and is more accurate by the determined frequency of the method or cycle.

Each zero crossing is respectively in the sampling of its both sides, and two sampled points of these zero crossing both sides that obtain living apart as the summit, namely form a vertical angle triangle with time value point corresponding to two sampled points and this two sampled point, and it consists of similar triangles.

According to the similar triangles theorem, have:

$\frac{t_{B1C1}}{t_{A1\mathrm{<figure-callout\; id="1"\; label="B"\; filenames="HDA00002135007100011.png"\; state="\{\{state\}\}">B</figure-callout>}1}}=\frac{\left|u_i\right|}{\left|u_{i-1}\right|}$

Then can get,

$\frac{t_{B1C1}}{t_{A1\mathrm{<figure-callout\; id="1"\; label="C"\; filenames="HDA00002135007100011.png"\; state="\{\{state\}\}">C</figure-callout>}1}}=\frac{\left|u_i\right|}{\left|u_{i-1}\right|+\left|u_i\right|}=\frac{\left|u_i\right|}{u_i-u_{i-1}}$

So have:

$t_{B1\mathrm{<figure-callout\; id="1"\; label="C"\; filenames="HDA00002135007100011.png"\; state="\{\{state\}\}">C</figure-callout>}1}=\frac{\left|u_i\right|}{u_i-u_{i-1}}\&times;t_{A1\mathrm{<figure-callout\; id="1"\; label="C"\; filenames="HDA00002135007100011.png"\; state="\{\{state\}\}">C</figure-callout>}1}=\frac{\left|u_i\right|}{u_i-u_{i-1}}\&times;T_S$

Here T _SBe sampling interval.

In like manner,

$t_{B2C2}=\frac{\left|u_j\right|}{u_j-u_{j-1}}\&times;T_S$

t _O1O2≈t _B1B2＝(j-i)×T _S+t _B1C1-t _B2C2

Thus, can try to achieve waveform frequency to be measured is:

$f=\frac{1}{t_{O1O2}}\&ap;\frac{1}{t_{B1B2}}=\frac{1}{(j-i)\&times;T_S+t_{B1C1}-t_{B2C2}}$

By above derivation as can be known, the raising of sampling rate helps the raising of frequency measurement accuracy.

For filtering noise disturbs, adopt the maximal possibility estimation theory to get a plurality of cycle measurement results and be averaged again, obtain the accurate estimated value of frequency.

According to the frequency values that accurately obtains, namely can carry out with this Measurement accuracy of various signals in the electric system, as carrying out the measurement of magnitude of voltage, current value according to the frequency of measuring, perhaps carry out power quality analysis etc.

Claims (7)

1. the measuring method of an electric system medium frequency is used for the frequency of the measured signal waveform of real-time follow-up electric system, it is characterized in that, the method comprises the steps:

(1) set up rectangular coordinate system according to waveform to be measured, wherein transverse axis is the time, and the longitudinal axis is signal value, and this waveform to be measured is symmetrical about transverse axis, and the intersection point of itself and transverse axis is theoretical zero crossing;

(2) waveform to be measured is carried out continuous sampling, find to be positioned at arbitrarily the first theoretical zero crossing O ₁Two sampled point u both sides and continuous _I-1And u _i, and be positioned at this first theoretical zero crossing O ₁After be adjacent and the second theoretical zero crossing O that the rate of change symbol is identical ₂Two sampled point u of both sides _J-1And u _j, wherein i, j are respectively the time series number of sampled point;

(3) determine approximate zero crossing according to sampled point

Straight line connects two sampled point u _I-1And u _i, with the intersection points B of itself and transverse axis ₁Be defined as the first approximate zero crossing, straight line connects two sampled point u _J-1And u _j, with the intersection points B of itself and transverse axis ₂Be defined as the second approximate zero crossing;

(4) according to above-mentioned definite approximate zero crossing, can measure the frequency f of signal waveform to be measured, namely

$f=\frac{1}{(j-i)\&times;T_s+t_{B_1{C}_1}-t_{B_2{C}_2}}$

In the formula, T _SBe sampling time interval,

Be the first approximate zero crossing B ₁With sampled point u _iThe time point C of correspondence on transverse axis ₁Between the time interval,

Be the second approximate zero crossing B ₂With sampled point u _jThe time point C of correspondence on transverse axis ₂Between the time interval.

2. the method for real-time measurement of electric system medium frequency according to claim 1 is characterized in that, described measured signal waveform is periodic waveform, such as sinusoidal waveform, cosine waveform or non-sinusoidal periodic waveform.

3. the method for real-time measurement of electric system medium frequency according to claim 1 and 2 is characterized in that, described measured signal is voltage signal or current signal, electric field signal or field signal.

4. one of according to claim 1-3 the method for real-time measurement of described electric system medium frequency is characterized in that, the described first theoretical zero crossing and the second theoretical zero crossing are positioned at the same period of waveform to be measured.

5. one of according to claim 1-4 the method for real-time measurement of described electric system medium frequency is characterized in that, the accurate measured value of described frequency can obtain by averaged after measuring the frequency values in a plurality of cycles.

6. one of according to claim 1-5 the method for real-time measurement of described electric system medium frequency is characterized in that, described waveform to be measured can be processed through digital filtering first, carry out sampled measurements after filtering direct current and the noise again.

7. the measuring method in voltage, electric current, electric field or magnetic field in the electric system realizes by the frequency of measuring respective waveforms, and wherein, described frequency adopts one of claim 1-6 described method to measure.

Images