Identification of a motor with multiple nonlinearities by improved genetic algorithm

Abstract

This paper presents a mathematical model that employs a new genetic algorithm for motor identification. Mechanical structures require precise motor information for high control performance. However, it is difficult to acquire accurate motor information and a genetic algorithm can be an adequate method to search unknown parameters using only angular position. The previous methods by using conventional genetic algorithms do not give the most optimal result since they cannot adjust the parameters with infinite precision. A new method is needed to identify uncertain motor information. This paper proposes a mathematical model that was searched by the newly proposed genetic algorithm. The induced motor model is verified through the real experiment. © Springer-Verlag Berlin Heidelberg 2005.