Online adaptive control via evolutionary algorithm

This paper presents an online adaptive control method for nonlinear systems using evolutionary algorithms (EAs). A general framework of constructing an EA-based online control system that processes problems continuously is first introduced with focus on handling external disturbance adaptively. In order to verify the method, a neural network controller and genetic algorithm, which have emerged as the most popular and successful techniques in computational intelligence, are employed as an example of the method. Two nonholonomic systems are considered, one being a wheeled mobile robot and the other rotary crane system. To emphasize the advantage of the proposed approach, strong and sudden disturbances are used in these tests. Simulations show that the method works effectively to control the perturbed systems to the desired states. © 2010 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.