Numerical differentiation based algorithm for power measurement

Abstract

Numerical differentiation based algorithm for measuring fundamental frequency, amplitude, phase, active and reactive power of power systems is presented in this paper. The numerical differentiation use central Lagrange interpolation with 7 points for signal representation and estimation computation. The fundamental frequency is first estimated at a maximal error of 0.001% with the sample or computation consequences of voltage or current signal. The frequency estimation needs high order differentiation of sinusoid wave and at most 2 cycles or 40 ms. The fundamental amplitude and phase are estimated using the fundamental information, which comes from the sample consequences and the high order differentiation of non-sinusoidal signals. The estimation accuracy of the amplitude and phase come up to 0.01%. Simulation analysis is accomplished with a voltage signal with 3 order harmonics in Matlab software.