Neural Assisted PID Controllers Design for Control Systems

P.C. Chou, A.T. Dzen, and S.C. Hsieh (Taiwan)


Soft computing, PID controller, and neural networks.


Soft computing [1], [2] is vastly discussed and used in academic as well as in industry these days. Neural networks (NN) can be used as controllers which required detail and careful design for most plants [3]. There exists no general rule for designing NN as a controller yet, because a fine tune is always necessary to match a specified plant [4]. In this paper, genetic algorithm (GA) [5], [6], [7], [8] approach is widely used for finding such a NN. This NN is a PID assistor in this paper. From the simulation results, better transient response can be expected if this assisted neural controller (NC) is selected.

Important Links:

Go Back