Investigation of the method of RMS measuring based on the digital filtration of the square of samples

Abstract

At the present time the most popular method to measure the root-mean-square value (RMS) is based on the summation of squares of samples over a time proportional to its period. The disadvantage of this method is associated with the presence of an additional error caused by the frequency deviation of the input signal and by the presence of additional harmonics. In this article we propose an approach based on the digital filtering of the square of samples. The analytical expressions for the estimating of the RMS measurement error of the proposed measurement method are obtained. The analysis of the error resulting from the frequency deviation of the input signal and the presence of harmonics is realized. Requirements for applied digital filter are defined. Various types of digital filters including moving average filters and cascade-integrator- comb filters are considered. The estimates of RMS measurement error with respect to the signals of real power networks are defined by simulation in programs Matlab and Simulink.