Autonomous motion controller design using genetic algorithms

This paper presents an evolutional compensator design for motion control systems using genetic algorithms (GA). The control system is composed of a robust 2-degrees-of-freedom (2DOF) compensator based on the coprime factorization description. The structural design and parameterization for the compensator essentially requires complicated theoretical and numerical procedures to realize the given control specifications. In this research, an autonomous design approach is proposed, where the optimal compensator structure and its parameters can be autonomously determined to satisfy the specified motion control performance by the optimization ability of GA, paying particular attention to the system robustness against load disturbances and variation of mechanical parameters. The effectiveness of the proposed compensator design has been verified by experimental results using prototype.