GA-based evolutionary identification algorithm for unknown structured mechatronic systems

Soft computing techniques, e.g., neural networks, fuzzy inference, evolutionary computation, and chaos theory, have been applied to a wide variety of control systems in industry because of their control capability and flexibility. They are also powerful to handle the complicated mechatronic systems with various nonlinearities which are difficult to model using mathematical formulas. In order to achieve the system identification of unknown structured mechatronic systems, This work presents a novel evolutionary algorithm using genetic algorithms (GAs), where the optimal mathematical structure of plant mechanisms and the combination of parameters can be autonomously determined by means of the optimization ability of the GA. The effectiveness of the proposed identification has been verified by experiments with comparative studies, using the typical mechanical systems with velocity controller.