A flicker measurement method based on feature reduction

Abstract

The application discloses a flicker measurement method based on feature reduction, and steps of the method comprise the following steps: extracting a power grid flicker envelope D(n) by using an improved Teager energy operator with two-point interval; analyzing and processing the obtained envelope D(n) by using a five-term Rife-Vincent window interpolation FFT method to obtain flicker modulation wave feature parameters: amplitude coefficient m i , frequency f i and initial phase angle θ i ; reducing the obtained feature parameters into a sinusoidal modulation signal, removing the amplitude coefficient m i from 4 / π and combining other parameters to reduce the sinusoidal modulation signal into a rectangular modulation signal; taking (1±10%) of the envelope signal D(n) as a threshold, comparing the number of data points of the reduced signal within the threshold, when the number of data points of one group is greater than N / 4, the greater one is the flicker modulation wave signal, and corresponding flicker parameters are output; if the number of data points of both groups is less than N / 4, no flicker occurs. The application provides a discrimination basis for flicker identification, overcomes the problem that a traditional flicker detection algorithm cannot effectively identify a flicker signal and is limited in use, and improves the detection precision of the flicker.

Description

Technical Field

[0001] This invention relates to the field of power systems, and specifically to a flicker measurement method based on feature reconstruction, which provides a basis for determining flicker modulation type. Background Technology

[0002] With the continuous grid connection of numerous nonlinear and impulsive loads, as well as new energy equipment, power grid fluctuations have led to increasingly severe flicker problems, causing serious impacts on industrial production and social life. Therefore, accurately obtaining relevant parameters of power system flicker is the primary prerequisite for flicker assessment and mitigation, and is of great significance for improving the power grid environment, evaluating the severity of flicker, and ensuring power grid security.

[0003] Most existing detection algorithms adopt the design concept of digital flicker meters, which involves first extracting the flicker signal envelope and then calculating the envelope parameters, such as using Hilbert transform to extract the flicker envelope. This method suffers from severe edge flying-wing problems and high computational complexity. Some researchers have proposed real-time flicker parameter detection methods based on FFT. However, in asynchronous sampling, spectral leakage and the picket-fence effect reduce the accuracy of these methods. Other methods utilize correlation coefficients to estimate flicker parameters, but grid disturbances and other interconnected load disturbances significantly impact this detection approach.

[0004] Most of these flicker parameter detection methods assume that the measurement is performed under the premise that the type of flicker modulation wave is known. However, the actual type of flicker modulation wave cannot be predicted in advance, which severely limits the application of these detection algorithms in practice. Summary of the Invention

[0005] The purpose of this invention is to address the limitation of existing technologies in practical applications by providing a flicker measurement method. Based on the flicker feature restoration method, this method can effectively solve the technical problem that existing technologies cannot accurately detect flicker parameters when flicker modulation waveforms change.

[0006] The technical solution of the present invention is as follows:

[0007] This invention provides a flicker measurement method based on feature reduction, comprising:

[0008] 1) Perform discrete sampling of the power grid signal at a fixed sampling rate; extract the flicker envelope of the sampled power grid signal using an improved Teager energy operator with two-point intervals. ;

[0009] 2) The obtained envelope is processed using the five-term Rife-Vincent window interpolation FFT method. After processing, the characteristic parameters of the flicker-modulated wave are obtained: amplitude coefficient. ,frequency and the initial phase angle of the envelope ;

[0010] 3) Restore the obtained characteristic parameters to a sinusoidal modulation signal, and then convert the amplitude coefficient... remove And combine with other parameters to reconstruct a rectangular modulation signal;

[0011] 4) Obtain the envelope signal Amplitude As a threshold, the number of points in both sets of restored signals within this threshold is counted. If the number of data points in any restored signal is greater than one-quarter of the number of sampling points, the restored modulation waveform corresponding to the largest number of data points within the threshold is taken as the actual flicker modulation waveform type, and the corresponding flicker parameters are output. If the number of points in both sets of restored signals within this threshold is less than one-quarter of the number of sampling points, it indicates that flicker has not occurred in the actual power grid.

[0012] Optionally, the improved energy operator envelope extraction method with two-point intervals is specifically expressed as follows:

[0013]

[0014] In the formula, To improve the energy value extracted by the energy operator, n is the nth sampling point. , , These are three sampling points spaced two points apart. The amplitude and frequency of the envelope signal can be quickly tracked through these three points, and they are independent of the initial phase angle of the signal.

[0015] Optionally, the flicker measurement method based on feature reduction is characterized in that the obtained envelope The expression is:

[0016]

[0017] In the formula For flicker signal, For grid voltage amplitude, This is the fundamental digital angular frequency of the power grid.

[0018] Optionally, the flicker measurement method based on feature reduction is characterized in that the five-term Rife-Vincent window interpolation FFT method is used to obtain the envelope. The functional expression for the characteristic parameters of the flicker-modulated wave obtained after processing is as follows:

[0019] The discrete expression for the five-term RV window function is:

[0020] In the formula, n is the nth sampling point, and N is the total number of sampling points. For the five-term RV window discrete expression, Multiplying by the five RV terms, the discrete expression of the windowed signal is obtained as follows:

[0021]

[0022] An FFT analysis was performed on the windowed flicker-modulated wave signal, and its discrete spectrum was obtained as follows:

[0023]

[0024] In the formula, For the number of discrete points, , For the number of modulation terms, For the initial phase angle of the envelope, For sampling frequency, This is the fundamental frequency of the power grid.

[0025] Optionally, the obtained flicker modulation amplitude coefficient ,frequency and the initial phase angle of the envelope for:

[0026]

[0027]

[0028]

[0029] In the formula, , The peak points detected are respectively The spectral lines with the largest and second largest amplitudes nearby and The amplitudes corresponding to the two spectral lines ,in The value range is [-0.5 0.5]. , Sampling frequency, The number of sampling points. For the discrete spectrum, the first Spectral lines, This is the fundamental frequency of the power grid.

[0030] Optionally, the flicker measurement method based on feature restoration is characterized in that the feature parameters are restored into a sinusoidal modulation signal, and then the amplitude coefficient is... remove And combined with other parameters, it is restored to a rectangular modulation signal, and the restored signals are as follows:

[0031]

[0032]

[0033] In the formula, n is the number of discrete points.

[0034] The optional flicker measurement method based on feature reduction is characterized in that the envelope signal is acquired. of Using the threshold as a criterion and counting the number of points within the threshold for the two sets of restored signals, the following conditions are met:

[0035]

[0036]

[0037] An optional flicker measurement method based on feature reconstruction is characterized in that, if the number of data points of the reconstructed signal is greater than... If the number of points is larger, the restored modulation waveform corresponding to that point is the actual flicker modulation waveform type, and the output of the corresponding flicker parameters is:

[0038] The number of points within the threshold is greater than At that time, the output flicker modulation amplitude coefficient ,frequency .

[0039] The number of points within the threshold is less than At that time, the output flicker modulation amplitude coefficient ,frequency .

[0040] As can be seen from the above technical solution of the present invention, the present invention has the following technical effects:

[0041] This invention is based on an improved Teager energy operator that utilizes a two-point interval, which can significantly improve the accuracy of flicker envelope extraction of power grid signals.

[0042] This invention provides a basis for classifying flicker modulation waveforms through a feature reconstruction measurement method, overcoming the problem that traditional flicker detection algorithms cannot effectively identify flicker signals and are therefore limited in use, thus improving the detection accuracy and reliability of flicker. Attached Figure Description

[0043] Figure 1 The flicker envelope extraction diagrams for rectangular modulation using traditional and improved energy operators are shown.

[0044] Figure 2 Spectral characteristic curves of different window functions;

[0045] Figure 3This is a flowchart illustrating the implementation of the present invention;

[0046] Figure 4 The simulation results of flicker signals of a single sine wave with different modulation frequencies are shown in the figure.

[0047] Figure 5 The simulation results of flicker signals with different modulation frequencies for a single rectangular wave are shown in the figure.

[0048] Figure 6 The graph shows the measurement results of sinusoidal-modulated flicker signals under different system frequency fluctuations.

[0049] Figure 7 The graph shows the measurement results of rectangular wave modulated flicker signals under different system frequency fluctuations. Detailed Implementation

[0050] The following will combine Figure 1 — Figure 7 The technical solution of the present invention will be further described in detail below.

[0051] The implementation steps include:

[0052] 1) Extracting the grid flicker envelope using the improved Teager energy operator with two-point spacing ;

[0053] 2) The obtained envelope is processed using the five-term Rife-Vincent window interpolation FFT method. Analysis and processing were performed to obtain the characteristic parameters of the flicker-modulated wave: amplitude coefficient. ,frequency and the initial phase angle of the envelope ;

[0054] 3) Restore the obtained characteristic parameters to a sinusoidal modulation signal, and then convert the amplitude coefficient... remove And combine with other parameters to reconstruct a rectangular modulation signal;

[0055] 4) Obtain the envelope signal of As a threshold, the number of points in both sets of restored signals within the threshold is counted. If the number of data points in any restored signal is greater than the threshold, then... If the number of points is larger, the restored modulation waveform corresponding to that point is the actual flicker modulation waveform type, and the corresponding flicker parameters are output; if the number of points in the restored signal within the threshold is less than [a certain value], then [the signal is considered flicker-modulated]. At that time, no flicker occurred.

[0056] Optionally, the detailed steps of step 1) include:

[0057] Flicker can be viewed as the result of modulating the mains frequency voltage by including one or more voltage fluctuation components as modulating waves. Its discrete mathematical expression is established as follows:

[0058]

[0059] In the formula, For sampling points; This refers to the voltage amplitude of the power grid. For the discrete components of flickering fluctuations; The initial phase angle of the fundamental wave; The fundamental digital angular frequency, The fundamental frequency, The sampling frequency.

[0060] Based on the traditional energy operator's sampling of three adjacent points, an improved energy operator envelope extraction method using two-point interval sampling is proposed. The specific expression is as follows:

[0061]

[0062] In the formula , , These are three sampling points spaced two points apart. The amplitude and frequency of the envelope signal can be quickly tracked through these three points, and they are independent of the initial phase angle of the signal.

[0063] For any discrete-time signal and The above energy operator satisfies

[0064]

[0065] Substituting the formula into the energy operator equation, we can see that...

[0066]

[0067] Due to the amplitude of the flicker modulated wave signal Typically smaller than the fundamental amplitude of the power grid. 10%, that is, the modulation coefficient Less than Therefore, in equation (4) Discarding smaller components and retaining only the main component simplifies it to:

[0068]

[0069] Finally, the improved energy operator flicker envelope component can be expressed as:

[0070]

[0071] Optionally, the detailed steps of step 2) include:

[0072] The discrete expression for the five-term RV window function is:

[0073]

[0074] The envelope signal in equation (6) is compared with the five RV functions. Multiplying them, we obtain the discrete expression of the windowed signal as follows:

[0075]

[0076] Analyzing the weighted flicker-modulated wave envelope signal, its discrete spectrum can be approximated as follows:

[0077]

[0078] In the formula, The number of discrete points, The term number of the flicker-modulated wave; ; The length to truncate the data; For the first The coefficients of the flicker-modulated wave; Let be the initial phase angle of the envelope. Let the peak point be... The spectral lines with the largest and second largest amplitudes nearby are respectively and The amplitudes corresponding to these two spectral lines are respectively and Introducing parameters and ,in The value range is [-0.5 0.5], which can be obtained from equation (9) through the definition of DTFT.

[0079]

[0080] Let the inverse function of equation (10) be: ,Depend on The parameters can be calculated. Then the flicker signal frequency The corrected formula is

[0081]

[0082] Correction coefficients for flicker-modulated waves Yes and The weighted average of the spectral lines is calculated using the following formula:

[0083]

[0084] Since the number of sampling points N is relatively large, therefore...

[0085]

[0086] From equation (9), we can obtain the envelope initial phase correction formula.

[0087]

[0088] In the formula, It was obtained through least squares fitting.

[0089] Therefore, equation (13) can also be expressed as

[0090]

[0091] Among them, take This represents the highest order of the polynomial fit.

[0092] Due to amplitude frequency response It is an even-symmetric function, therefore It is an even function, that is

[0093]

[0094] Therefore, we can obtain the first... The coefficient of the flicker modulated wave signal amplitude is modified as follows:

[0095]

[0096] Then rewrite equation (10) as And define its inverse function as .Pick exist A set of values ​​within the range, the corresponding values ​​are obtained from equation (10). The value is then inversely fitted to obtain the polynomial. The coefficients are then...

[0097]

[0098] Therefore, we can obtain the first... The frequency correction of the flicker modulated wave signal is as follows:

[0099]

[0100] The envelope initial phase angle is defined by DTFT. It can be represented as

[0101]

[0102] Optionally, the detailed steps of step 3) include:

[0103] The characteristic parameters are restored to a sinusoidal modulation signal, and then the amplitude coefficient is... remove And combined with other parameters, it is restored to a rectangular modulation signal, and the restored signals are as follows:

[0104]

[0105]

[0106] In the formula, n is the number of discrete points.

[0107] Optionally, the detailed steps of step 4) include:

[0108] Obtain the envelope signal of Using the threshold as a criterion and counting the number of points within the threshold for the two sets of restored signals, the following conditions are met:

[0109]

[0110]

[0111] If the number of data points of the restored signal is greater than If the number of points is larger, the restored modulation waveform corresponding to that point is the actual flicker modulation waveform type, and the output of the corresponding flicker parameters is:

[0112] The number of points within the threshold is greater than At that time, the output flicker modulation amplitude coefficient ,frequency .

[0113] The number of points within the threshold is less than At that time, the output flicker modulation amplitude coefficient ,frequency .

[0114] The effects of this invention can be obtained through the following simulation experiments, as detailed below:

[0115] 1. Simulation of a single-phase modulated flicker signal

[0116] To verify the accuracy and reliability of the proposed algorithm, this paper conducts simulation experiments in MATLAB, setting the sampling frequency... Number of sampling points , grid fundamental voltage and frequency They are respectively , And respectively set a sinusoidal wave modulated flicker signal and a rectangular wave modulated flicker signal, wherein the number of modulation wave terms is... ,coefficient exist Random selection ( (indicates nominal value) and frequency exist Range according to steps Simulations were performed to illustrate the changes. The results of the sinusoidal wave modulated flicker signal and the rectangular wave modulated flicker signal are shown below. Figure 4 , Figure 5 As shown.

[0117] Depend on Figure 4 As can be seen, the algorithm presented in this paper can accurately identify the type of flicker-modulated wave and obtain its parameters. From Figure 4 As can be seen, the sinusoidal modulated flicker signal exhibits high accuracy in extraction across different modulation coefficients and frequencies, with relative errors in the measurement results all below 0.02%. Figure 5 It can be seen that although the relative error of the rectangular wave modulated flicker signal is relatively large at 25Hz and 35Hz, it still meets the requirements of GB / T12326-2008.

[0118] 2. Simulation of multi-modulation flicker signals

[0119] To verify the accuracy and reliability of the proposed algorithm, a simulation experiment was conducted in MATLAB, with the sampling frequency set as follows: Number of sampling points , grid fundamental voltage and frequency They are respectively , And respectively set a sinusoidal wave modulated flicker signal and a rectangular wave modulated flicker signal, wherein the number of modulation wave terms is... ;coefficient Take values ​​of [0.03, 0.07, 0.1] respectively; frequency The Hz values ​​were set to [4, 10, 18]. The results of the sinusoidal wave modulated flicker signal and the rectangular wave modulated flicker signal are shown in Table 1.

[0120] As shown in Table 1, the method proposed in this invention can still accurately identify the modulation wave type and obtain relevant parameters for multi-modulation flicker signals. The maximum relative error for sinusoidal wave modulated flicker signals is 8.9E-04%, and the maximum relative error for rectangular wave modulated signals is 0.1783%, indicating that the algorithm of this invention also has high accuracy in detecting flicker parameters for multiple modulation wave types.

[0121] 3. Simulations for the influence of different fundamental frequencies

[0122] The actual power grid frequency fluctuates, which directly affects the accuracy of flicker parameter detection. Therefore, setting the modulation coefficient is crucial. Modulation wave number Fundamental frequency exist Changes, step size The specific simulation results are as follows: Figure 6 , Figure 7 As shown.

[0123] Depend on Figure 6 As can be seen, the algorithm presented in this paper can accurately identify the type of flicker-modulated wave and obtain its parameters. From Figure 6 As can be seen, the sinusoidal modulated flicker signal exhibits high accuracy in extraction across different modulation coefficients and frequencies, with relative errors in the measurement results all below 0.2%. Figure 7 It can be seen that although the relative error of the rectangular wave modulated flicker signal is relatively large at a few frequencies, it still meets the requirements of GB / T12326-2008.

[0124] 4. Simulation of measurements under the influence of 50dB Gaussian white noise

[0125] To verify the measurement results of the proposed detection method under noise conditions, a simulation test was conducted on a flicker signal superimposed with 50dB Gaussian white noise. Considering the randomness of the noise, 1000 measurements were performed for each flicker signal, and the average error was taken. The specific simulation test results are shown in Table 2.

[0126] As shown in Table 2, despite the influence of noise, the algorithm proposed in this invention can still accurately detect flicker signals. The relative error of the coefficients obtained from sinusoidal wave modulated flicker signals does not exceed 0.2699%, and the relative error of the frequency does not exceed 7.4e-04%. For rectangular wave modulated flicker signals, the relative error of the coefficients does not exceed 0.7462%, and the relative error of the frequency does not exceed -0.0520. Therefore, the proposed algorithm not only accurately measures flicker parameters but also exhibits good robustness.

[0127] 5. Verification of the application of this invention in actual flicker measurement.

[0128] To verify the accuracy and effectiveness of the algorithm presented in this paper for detecting flicker parameters in a real power grid, a flicker testing hardware platform based on STMicroelectronics' STM32F407 ARM processor and Analog-to-digital converter AD7606 was constructed. The overall structure of the flicker testing hardware platform is as follows: Figure 6 As shown, a signal generator (Fluke 6100A) generates a flicker signal for testing. This signal, after passing through a signal conditioning circuit, is sent to an analog-to-digital converter (AD7606) for data acquisition. The acquired data is transmitted to an STM32F407 via serial communication (SPI). The STM32 performs flicker parameter detection and analysis, and the analysis results are displayed on an LCD. An Agilent DS01102B oscilloscope is used for real-time monitoring of the generated flicker signal.

[0129] The flicker parameter results obtained from the actual hardware test platform are shown in Table 3.

[0130] As shown in Table 3, the algorithm presented in this paper can accurately identify the specific modulation wave type of flicker. The relative error of the amplitude of the sinusoidal wave modulation flicker obtained from actual tests does not exceed 0.7321%, and the relative error of the frequency does not exceed 0.1870%. The amplitude error of the rectangular wave modulation flicker does not exceed 0.7886%, and the frequency error does not exceed 0.0635%, both of which meet the flicker detection error requirements specified in GB / T12326-2008 standard.

[0131] As can be seen from the above, the flicker measurement method based on feature reconstruction proposed in this invention can accurately identify flicker modulation wave types and accurately detect related parameters. It effectively solves the problem that traditional flicker detection methods cannot identify flicker signals of different modulation wave types, thus limiting their application. Furthermore, the algorithm has high accuracy and good anti-interference capabilities, providing an effective basis for classifying flicker modulation waveforms.

[0132] The above description represents preferred embodiments of the present invention, but these embodiments do not limit the scope of the invention. Any equivalent changes or modifications made without departing from the spirit of the present invention are also within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the claims of this application.

Claims

1. A flicker measurement method based on feature reduction, characterized in that, The flicker measurement method includes: 1) Extracting the grid flicker envelope using the improved Teager energy operator with two-point spacing ; 2) The obtained envelope is processed using the five-term Rife-Vincent window interpolation FFT method. Analysis and processing were performed to obtain the characteristic parameters of the flicker-modulated wave: amplitude coefficient. ,frequency and the initial phase angle of the envelope ; 3) Restore the obtained characteristic parameters to a sinusoidal modulation signal, and then convert the amplitude coefficient... remove And combine with other parameters to reconstruct a rectangular modulation signal; 4) Obtain the envelope signal of As a threshold, the number of points of the two sets of restored signals within the threshold is counted. If the number of data points of the restored signal is greater than one-quarter of the number of sampling points, the restored modulation waveform corresponding to the larger number of points is the actual flicker modulation waveform type and the corresponding flicker parameters are output. If the number of points of the restored signal within the threshold is less than one-quarter of the number of sampling points, then no flicker has occurred.

2. The flicker measurement method based on feature reduction according to claim 1, characterized in that, The improved energy operator envelope extraction method with two-point intervals is specifically expressed as follows: ， In the formula, , , These are three sampling points spaced two points apart. The amplitude and frequency of the envelope signal can be quickly tracked through these three points, and they are independent of the initial phase angle of the signal.

3. The flicker measurement method based on characteristic parameter restoration according to claim 2, characterized in that, The obtained envelope The expression is: ， In the formula For flicker signal, For grid voltage amplitude, This is the fundamental digital angular frequency of the power grid.

4. The flicker measurement method based on feature reduction according to claim 1, characterized in that, The five Rife-Vincent window interpolation FFT methods are used to obtain the envelope. The functional expression for the characteristic parameters of the flicker-modulated wave obtained after processing is as follows: The discrete expression for the five-term RV window function is: ， Will Multiplying by the five RV terms, the discrete expression of the windowed signal is obtained as follows: ， The discrete spectrum of the windowed flicker-modulated wave signal obtained by FFT analysis is as follows: ， In the formula, The number of discrete points, , The number of modulation terms, Sampling frequency, This is the fundamental frequency of the power grid.

5. The flicker measurement method based on feature reduction according to claim 4, characterized in that, The obtained flicker modulation amplitude coefficient ,frequency and the initial phase angle of the envelope for: ， ， ， In the formula, , The peak points detected are respectively The spectral lines with the largest and second largest amplitudes nearby and The amplitudes corresponding to the two spectral lines ,in The value range is [-0.5 0.5]. , Sampling frequency, The number of sampling points. For the discrete spectrum, the first Spectral lines, This is the fundamental frequency of the power grid.

6. The flicker measurement method based on feature reduction according to claim 1, characterized in that, The process involves restoring the characteristic parameters to a sinusoidal modulation signal, and then converting the amplitude coefficient... remove And combined with other parameters, it is restored to a rectangular modulation signal, and the restored signals are as follows: ， ， In the formula, n is the number of discrete points.

7. The flicker measurement method based on feature reduction according to claim 1, characterized in that, The aforementioned envelope signal acquisition of Using the threshold as a set of values, and counting the number of points within the threshold for the two sets of restored signals, the following conditions are met: ， 。 8. The flicker measurement method based on feature reduction according to claim 1, characterized in that, If the number of data points of the restored signal is greater than The restored modulation waveform corresponding to the larger number of points is the actual flicker modulation waveform type, and the output of the corresponding flicker parameters is: The number of points within the threshold is greater than At that time, the output flicker modulation amplitude coefficient ,frequency and the initial phase angle of the envelope ; The number of points within the threshold is less than At that time, the output flicker modulation amplitude coefficient ,frequency and the initial phase angle of the envelope .

Images