Chaos optimization neural network control for the stability of double inverted pendulum

Aiming at the double inverted pendulum system, a neural controller based on chaos optimization method was designed. To solve the local minimum problem, the disadvantage of BP neural network, an “two-stage” chaos optimization algorithm which is easy to get the global optimization was adopted. By the use of BP algorithm, rough breadth chaos search and elaborate depth chaos search, the global optimization weight values of neural network were obtained, in which the knowledge of control is implied. Because of the capability of non-linear mapping and generalization, the rule-based neural controller succeeded in the equilibrium control of the double inverted pendulum system. Physical control experiments show that this optimization strategy is feasible and effective.