Robust estimation of load performance of DC motor using genetic algorithm

This paper presents a novel approach to estimate the load performance curves of DC motors whose equations are represented as a function of the torque based on a steady-state model with constraints. Since a simultaneous optimization of the curves forms a multi-objective optimization problem (MOP), we apply an optimal curve fitting method based on a real-coded genetic algorithm (RGA). In the method, we introduce a normalized ratio of errors to solve the MOP without the use of weighting factors and the nominal parameters to automatically determine the searching bounds of the curve parameters. Compared to the conventional least square fitting methods, the proposed scheme provides robust and accurate estimation characteristics even when fewer measurements with a small range of torque loading are taken and used for a data fitting.