Multiharmonic Waveform Fitting of Periodic Signals using Genetic Algorithms

A genetic algorithm is used to estimate the parameters of periodic signals in multiharmonic waveform fitting. This algorithm is used to find the fundamental frequency which can then be used in a multiple linear least-squares (LS) waveform fitting algorithm. This procedure is applied to different multiharmonic signals with very good results. The main advantage of genetic algorithms when compared with traditional LS minimization methods is that genetic algorithms can recover whenever the operation point in the iteration falls on a local maximum or minimum. This method is applied to experimentally acquired signals used for impedance measurements.