CN104569581B - Multi-level set and single-cycle estimation method of power grid frequency measuring - Google Patents

Abstract

The present invention provides a multi-level-set single-period estimation method for power grid frequency measurement: first, perform low-pass filter processing on the input signal, and then perform sampling; then set two or more thresholds to intersect with the sinusoidal period signal level, and determine Each threshold has the same adjacent intersection points with the same slope as the 2 tangent lines of the sinusoidal curve and the signal sampling points adjacent to these two intersection points. If the two intersection points of a certain threshold coincide with the signal sampling points, then directly calculate its fundamental frequency, otherwise Substitute the signal sampling points near the intersection above into the Lagrange interpolation formula to calculate the estimated value of the fundamental frequency corresponding to the threshold; finally, determine the multi-level set basis obtained by each threshold according to the absolute value of the slope of the tangent line at the intersection of each threshold and the sinusoidal signal. The weight of the estimated value of the fundamental frequency is obtained by using the weighted average method to obtain the final estimated value of the fundamental frequency. The invention uses multiple thresholds to realize multi-level set single-cycle frequency estimation, and has the advantages of good real-time performance and high precision.

Description

A multi-level-set single-period estimation method for power grid frequency measurement

technical field

The invention belongs to the field of electric power, and in particular relates to a method for measuring the frequency of a grid.

Background technique

In the existing fundamental wave frequency measurement method, the whole cycle sampling point counting method is often used. In this method, the cycle of the fundamental wave is determined first. the frequency of the wave. When the sampling period is small, the accuracy of the measured frequency is very poor, and when the sampling period is long, it cannot meet the real-time requirements. In addition, this method does not filter out high-frequency harmonic signals when measuring frequency. When the harmonic interference is severe, it will affect the determination of the entire fundamental wave cycle, which will eventually lead to inaccurate measured frequencies.

Contents of the invention

The purpose of the present invention is to overcome the shortcomings of the above grid frequency measurement method, and propose a multi-level set single-cycle estimation method for grid frequency measurement. The problem of poor real-time performance.

The present invention proposes a multi-level-set single-period estimation method for grid frequency measurement, said method comprising the following steps:

Step 1: Perform low-pass filtering processing on the input signal with a low-pass filter; eliminate the interference of high-frequency harmonics;

Step 2: Select the appropriate sampling frequency f _s and sampling length N, and sample the signal at equal intervals. The sampling frequency f _s should not be lower than twice the frequency of the highest harmonic component contained in the signal, and obtain the signal discrete sampling sequence x( t _n ), n=0,1,2,...,N-1;

Step 3: Find the maximum value MAX and the minimum value MIN in the discrete sampling sequence x(t _n ), and then set m thresholds Y _i , where m is an integer greater than or equal to 2, i=1, 2,..., m;

Step 4: Determine each threshold and the adjacent intersection points with the same tangent slope of the sinusoidal curve and the signal sampling points adjacent to these two intersection points. The time interval between the two adjacent intersection points with the same tangent slope is a basis of the signal wave period; if the two intersection points of a certain threshold coincide with the signal sampling point, then directly calculate the estimated fundamental frequency f _i ; if the two intersection points of a certain threshold do not coincide with the signal sampling point, then the threshold and the sine The signal sampling points near the above intersection point of the signal are substituted into the Lagrange interpolation formula to calculate the estimated value of the fundamental frequency corresponding to the threshold Y _i ; after all the thresholds are calculated, the obtained m fundamental frequency estimates are multi-level set fundamental frequency estimates value f _i ;

Step 5: Perform weighted average on the multi-level set fundamental frequency estimate f _i obtained in step 4, and calculate the final fundamental frequency estimate f, where the weight q _i is based on the tangent of the sinusoidal curve at the intersection point of each threshold and the sinusoidal curve The absolute value of the slope |K _i | is determined, the greater the absolute value of the slope, the smaller the weight, and the final estimated fundamental frequency f=q ₁ f ₁ +q ₂ f ₂ +...+q _m f _m .

In the method, in step 3, the threshold Y _i should be less than the maximum value in the discrete sampling sequence x(t _n ) and greater than the minimum value in the discrete sampling sequence x(t _n ), and to ensure that the set threshold Within a reasonable range, eliminate the interference caused by other uncertain factors to meet: 0.9MIN+0.1MAX<Y _i <0.9MAX+0.1MIN;

In the described method, in step 4, if 2 intersection points of a certain threshold do not coincide with the signal sampling points, then the signal sampling points near the above-mentioned intersection points of the threshold and the sinusoidal signal are substituted into the Lagrange interpolation formula to calculate the corresponding threshold Y _i The estimated value of the fundamental frequency, in order to simplify the calculation, the discrete data (t, x(t)) of the sampled signal near the intersection of the threshold and the sinusoidal signal is substituted into the quadratic Lagrange interpolation formula, and the fundamental frequency corresponding to the threshold Y _i is obtained through calculation Estimated value f _i .

In the described method, in step five _, the weight q _i is determined according to the slope absolute value |K _i |

Beneficial effects: the invention adopts multi-level sets to realize power grid frequency measurement, which overcomes the defect that the previous single-threshold method is easily disturbed by noise, and the invention can realize frequency measurement in a single cycle, and has the characteristics of high precision and good real-time performance.

Description of drawings

Fig. 1 is a functional block diagram of the process flow of the present invention;

Fig. 2 is a schematic diagram of single threshold frequency estimation in the present invention;

Fig. 3 is the schematic diagram of the frequency estimation in which the intersection point and the sampling point coincide in the present invention;

Fig. 4 is the schematic diagram of the frequency estimation where the intersection point and the sampling point do not coincide in the present invention;

Fig. 5 is a principle diagram of multi-threshold frequency estimation in the present invention.

detailed description

The invention proposes a multi-level set single-period estimation method for grid frequency measurement. The present invention will be described in further detail. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

1. In the embodiment of the present invention, the input signal is first low-pass filtered with a low-pass filter, and then the signal is sampled at equal intervals with a sampling frequency f _s =1000 Hz to find the maximum value MAX and the minimum value MIN;

The model of the fundamental sinusoidal periodic signal is:

x(t)＝A sin(2πft+θ)+B (1)

In the formula, A represents the amplitude of the sinusoidal periodic signal, f represents the frequency of the sinusoidal periodic signal, θ represents the initial phase of the sinusoidal periodic signal, and B represents the DC component of the sinusoidal periodic signal. As shown in Figure 2, the time interval between two adjacent intersection points where the sinusoidal periodic signal x(t) intersects with a certain threshold b and the slope of the tangent line of the sinusoidal curve is the same is one fundamental period of the signal. Therefore, through these two intersection points, an estimated value of the fundamental frequency of the signal can be obtained.

2. The periodical analog signal shown in formula (1) is sampled at equal intervals with the sampling frequency f _s . In general, in the actual process, the intersection point of the threshold b and the sinusoidal signal is generally difficult to coincide with the signal sampling point, so it is divided into two Cases:

① As shown in Figure 3, the above intersection point of the threshold b coincides with the signal sampling point, and the estimated value of the fundamental frequency corresponding to this threshold should be directly calculated:

where f _g is the estimated value of the fundamental frequency corresponding to the threshold b;

②As shown in Figure 4, the intersection point of the threshold _b does not coincide with the signal sampling point, and the sampling points [t _k-2 , x(t _k-2 )], [t _{k -1} , x(t _k-1 )], [t _k , x(t _k )], [t _k+1 , x(t _k+1 )] and [t _k+2 , x(t _k+2 )]. Using the Lagrange interpolation method, the 4th degree interpolation polynomial L ₄ (t) can be obtained as:

Substitute (t _b ,b) into (3) to get:

L ₄ (t _b )=b (4)

Solving formula (4) to get t _b , similarly, we can get t _b ′, as follows:

where f _g is the estimated value of the fundamental frequency corresponding to the threshold b; so far, the estimation of the fundamental frequency corresponding to the threshold b has been completed.

3. For the convenience of description, the following only explains the intersection of the five thresholds and the sinusoidal periodic signal:

Set 5 thresholds Y _i that intersect with the sinusoidal periodic sampling signal, and make them within a reasonable range, 0.9MIN+0.1MAX<Y _i <0.9MAX+0.1MIN, where i=1,2,3,4,5 . The estimated value of the fundamental frequency corresponding to each threshold can be obtained from the above principles, and the estimated value of the fundamental frequency corresponding to the threshold Y ₁ is f ₁ ; the estimated value of the fundamental frequency corresponding to the threshold Y ₂ is f ₂ ; The estimated fundamental frequency corresponding to the threshold Y ₃ is f ₃ ; the estimated fundamental frequency corresponding to the threshold Y ₄ is f ₄ ; the estimated fundamental frequency corresponding to the threshold Y ₅ is f ₅ . According to the magnitude of the slope absolute value |K _i | of the tangent line at the sinusoidal curve at the intersection point of each threshold Y _i and the sinusoidal period sampling signal, the value of K _i is determined by the 4th degree interpolation polynomial L ₄ (t) of the above intersection point of Y _i and the sinusoidal curve The derivative of is found, that is:

Weight the estimated fundamental frequency f _i corresponding to each threshold, the larger the absolute value, the smaller the weight, and the weight corresponding to f _i is i = 1, 2, 3, 4, 5; then use the weighted average method to calculate the final frequency estimate:

where f is the final fundamental frequency estimate.

So far, the multi-level set single-period estimation of grid frequency measurement has been completed. This method is simple to calculate, and the fundamental frequency of the power grid can be measured through the adjacent intersection point whose threshold value is the same as the slope of the tangent line of the sampled signal sinusoidal curve, and uses the 4th Lagrange interpolation method and weighting method, which has very good real-time performance and accuracy sex.

Claims (4)

1. the multilevel collection monocycle method of estimation of a kind of grid frequency measurement, it is characterised in that methods described includes following steps Suddenly：

Step one：Low-pass filtering treatment is carried out to input signal with low pass filter, the interference of high-frequency harmonic is eliminated；

Step 2：Select suitable sample frequency f_sWith sampling length N, equal interval sampling, sample frequency f are carried out to signal_sShould not Less than 2 times of the frequency of highest harmonic componentss contained by signal, signal discrete sample sequence x (t are obtained_n), n=0,1,2 ..., N- 1；

Step 3：Find out discrete sampling sequence x (t_n) in maximum MAX and minimum value MIN, then arrange m threshold value Y_i, m is big In the integer equal to 2, i=1,2 ..., m；

Step 4：Determine each threshold value and sinusoidal 2 tangent slope identical adjoining nodes and close on this 2 intersection points Signal sampling point, the time interval of 2 tangent slope identical adjoining nodes are 1 primitive period of signal, if certain 2 intersection points of threshold value are overlapped with signal sampling point, then directly calculate fundamental frequency estimated value f_iIf, 2 of certain threshold value Intersection point is not overlapped with signal sampling point, then the signal sampling point by the threshold value with the above-mentioned near intersections of sinusoidal signal is substituted into Lagrange formula for interpolations, calculate threshold value Y_iCorresponding fundamental frequency estimated value, after all threshold calculations are finished, the m of acquisition Individual fundamental frequency estimated value is multilevel collection fundamental frequency estimated value f_i；

Step 5：To multilevel collection fundamental frequency estimated value f is obtained in step 4_iIt is weighted average, the final fundamental wave frequency of calculating Rate estimated value f, wherein weights q_iAccording to the sine curve tangent slope absolute value of each threshold value and sinusoidal intersection point | K_i| It is determined that, slope absolute value is bigger, and weights are less, final fundamental frequency estimated value f=q₁f₁+q₂f₂+…+q_mf_m。

2. method according to claim 1, it is characterised in that in step 3, described threshold value Y_iDiscrete sampling should be less than Sequence x (t_n) in maximum and be more than discrete sampling sequence x (t_n) in minima, and for ensure set by threshold value rational In the range of, eliminate the interference that other uncertain factors cause so as to meet 0.9MIN+0.1MAX<Y_i<0.9MAX+0.1MIN。

3. method according to claim 1, it is characterised in that in step 4, if 2 intersection points of certain threshold value not with letter Number sampled point overlaps, then will the threshold value to substitute into Lagrange interpolation with the signal sampling point of the above-mentioned near intersections of sinusoidal signal public Formula, calculates threshold value Y_iCorresponding fundamental frequency estimated value, is simplified operation, by adopting for the threshold value and sinusoidal signal near intersections Sample signal discrete data (t, x (t)) substitute into 4 Lagrange formula for interpolations, through being calculated threshold value Y_iCorresponding fundamental wave Frequency estimation f_i。

4. method according to claim 1, it is characterised in that in step 5, weights q_iAccording to each threshold value With the sine curve tangent slope absolute value of sinusoidal intersection point | K_i| it is determined that, f_iCorresponding weights are：