Research on Voltage Flicker Measurement Based on LabWindows /CVI

0 Introduction

The load with impact (rapid change) power in the power system will cause the voltage fluctuation and flicker of the power grid, causing many electrical equipment to fail to work properly, seriously affecting the power quality of the power grid. Therefore, the accurate measurement of voltage fluctuation and flicker is becoming more and more important.

The International Electrotechnical Commission (IEC) has given international unified standards for flicker measurement and evaluation, but IEC has not given specific implementation methods. This paper uses the virtual instrument development platform LabWindows/CVI to develop a flicker measurement module based on the IEC flicker measurement principle. The measurement accuracy is verified to fully meet the IEC standard.

1 IEC flicker measurement principle

The principle block diagram of the flicker meter recommended by IEC is shown in Figure 1. The input quantity is a voltage signal. After being processed by the filters in blocks 1 to 4, the output signal instantaneous flicker visual sensitivity S(t) is obtained. This signal reflects the impact of light flickering caused by voltage fluctuations on human vision. After online or offline statistical evaluation of S(t), the short-term flicker value Pst or the long-term flicker value Plt can be obtained.

2 Determination of filter parameters in IEC flicker meter

The function of block 1 in the flicker meter block diagram recommended by IEC is to square the input signal and send it to block 2, where a 0.05 Hz to 35 Hz filter is used to filter out the DC component and the frequency components above the power frequency. The filter is realized by a 0.05 Hz high-pass filter and a 35 Hz low-pass filter in series.

The transfer function of a 0.05 Hz high-pass filter is:

The transfer function of the sensitivity weighting filter in box 3 is:

The function of the first-order low-pass filter in box 4 is to simulate the reflection and memory effect of human brain nerves on vision, and its transfer function is:

In formula (4), τ=300 ms.

To realize the above filter in a digital system, it needs to be digitized, usually by impulse response invariance method and bilinear transformation method. In this design, bilinear transformation method is adopted, which can be realized by using the transformation function in MATLAB: [NUMd, DENd] = BILINEAR (NUM, DEN, Fs). For example, the MATLAB program to convert a 35 Hz Butterworth low-pass filter from an analog filter to a digital filter is:

Where n and d are the numerator and denominator coefficients of the analog filter transfer function, respectively. The sampling frequency used in this design is 3,200 Hz. The Nd and Dd obtained after running are the numerator and denominator coefficients of the digital filter.

3 IEC Flicker Meter Hardware and Software Implementation

The hardware structure of the voltage flicker measurement module is shown in Figure 2. After the grid voltage passes through the transformer board, the high voltage is reduced to a low voltage signal between -5 V and +5 V, and then sent to the acquisition card for AD conversion. The converted digital quantity is sent to the industrial computer for analysis and calculation, and finally the flicker value is obtained.

According to the national standard, short-time flicker calculation requires uninterrupted measurement for 10 minutes, that is, 600 seconds. Since the convergence of the filter requires a certain amount of time, as shown in Figure 3, this design requires uninterrupted measurement for 650 seconds each time a short-time flicker is calculated. When calculating the flicker value, the first 50 seconds of data are discarded to ensure that the filter has converged to a steady-state value at this time. According to MATLAB simulation, the above-mentioned filter has entered a steady state at 30 seconds. In order to ensure the accuracy of the calculation, this design uses data after 50 seconds.

The filter used in flicker calculation is implemented by using the filter function IIRFiltering() provided by LabWindows/CVI. After the numerator and denominator polynomial coefficients of the digital filter system function are given, the IIRFiltering() function can perform IIR filtering on the input data. The function for processing data is shown below. This function processes a packet of collected data, and after processing, the visual sensitivity curve S(t) can be obtained. In this design, the sampling frequency is 3,200 Hz, and a packet of data has 4 cycles, so there are 256 data in each packet of data. Among them, LPa and LPb are the numerator and denominator coefficients of the 35 Hz low-pass filter system function, and LPx and LPy are the initial conditions after the filter converges.

After obtaining the visual sensitivity curve S(t), according to the method recommended by IEC, it is necessary to perform equal-periodic sampling, graded timing, and calculate the cumulative probability function CPF. According to CPF, the short-term flicker value Pst and the long-term flicker value Plt are finally obtained. The calculation formula for Pst is:

According to IEC regulations, after continuous measurement for 2 hours and obtaining 6 short-time flicker values, the long-time flicker value Plt can be calculated.

4. Verification Data

IEC provides the flicker measurement model and the calibration standard of the flicker tester, as shown in Table 1. When the rectangular amplitude modulation wave with the specified value in the table is input, Pst = "14-0". 05 is required to be measured. In order to verify this design, the calibration method given by IEC was used for testing. The third column in Table 1 is the short-time flicker value measured under the condition of sampling frequency fs = 3 200 Hz. It can be seen that the error is within the range specified by IEC and meets the IEC standard.

5 Conclusion

In this paper, according to the flicker measurement method recommended by IEC and combined with the design concept of virtual instrument, a voltage flicker measurement module is developed on the LabWindows/CVI development platform. As a part of the power quality analysis system, the calculation accuracy is fully in line with the IEC standard after simulation and practical application verification.