Harmonic frequency estimation under nonsinusoidal conditions

Abstract

Accurate estimation of power system fundamental frequency is a necessity to check the state of health of the power index. This paper presents a hybrid technique for harmonic frequency estimation of non-sinusoidal signals of power systems with high order harmonics using central numerical differentiation and digital FIR filter algorithm. A digital low-pass FIR filter is used to obtain the discrete values of the fundamental sinusoidal component, and a set of digital bandpass FIR filter is applied to compute the discrete values of the harmonic component. Using 6 points numerical differentiation, these discrete values are used for fundamental and harmonic frequency estimation. For a signal with 5 harmonics, the technique requires at most 2 cycles for the first time of estimation computation, and requires at most 1 cycle for the latter estimation. Comparing with other existing techniques, the proposed algorithm is characteristic of high accuracy and much less time. With satisfactory results, a study example is given to illustrate the proposed algorithm in Matlab.