An improved adaptive PID controller algorithm

For many nonlinear and time varying industrial processes with strong couplings, existing improved PID controllers still produce unsatisfactory dynamic and steady-state perfomances. A new multivariate adaptive PID controller is proposed in this paper which combines the Fuzzy control and RBF neural networks. The proposed controller is able to improve the control performance meanwhile the parameters can be tuned online. The new controller can further improve the stability of the nonlinear system. Simulation results show that the proposed algorithm converges faster and is more robust, and the dynamic and static performances are also significantly inproved.