Performance Comparison between Complex Non-Linear Least Squares and Genetic Algorithms in Impedance Circuit Parameter Estimation

Abstract

Electric circuit parameter estimation from its impedance frequency response is becoming increasingly important due to the wide range of applications that it finds both in science and industry. This paper presents a comparison of the performance of a genetic algorithm and the complex non-linear least squares implemented in the LEVM/LEVMW software in the estimation of impedance circuit parameters. The comparison is focused on the standard deviation of the estimated values, convergence rate and execution time of each algorithm.