Neuro-PID Control for Stabilization of Inverted Pendulums
by
Sigeru Omatu
Osaka Prefecture University
Coauthors: Toru Fujinaka, Michifumi Yoshioka, and Sigeru Omatu(Osaka Prefecture University)

Computational intelligence approaches like fuzzy theory, neural network, genetic algorithm, artificial life, etc. have been well-developed and applied to many real control problems in an efficient way[1]-[4]. In real control processes, the PID controller has been used as the major control methods and the operators select suitable PID gains from time to time based on their experience and knowledge about the plant dynamics. However, it is required to find a suitable PID gains automatically. The main reason is that its structure is simple and PID controllers are robust with noise and parameter variations. To use the PID controller, we must tune the PID gains which have been determined by trial and error based on experience and knowledge of experts. This paper is concerned with a new architecture of a self-tuning neuro-PID control system and its application to stabilization of an inverted pendulum. A single-input multi-output system is considered to control the inverted pendulum by using the PID controller. The PID gains are tuned by using two kinds of neural networks. The simulation results show effectiveness of the proposed approach.

Date received: November 13, 2001

Copyright © 2001 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cago-07.