Application of Genetic Algorithms in Sinewave Fitting

Abstract

In this paper, two genetic algorithm variants are applied to sine fitting of acquired waveforms, to solve the convergence problems that are usually associated with sine fitting algorithms. The genetic algorithms are used to find the sinewaves’ frequency, amplitude, phase and DC component that best fit, in a least-squares sense, the acquired samples. The robustness of these algorithms to convergence problems is also demonstrated. The algorithm is then successfully applied to experimentally acquired waveforms.