Using a neural network as a feedback controller

Author

Cheung, John Y. ; Mulholland, Robert J.

Author_Institution

Sch. of Electr. Eng. & Comput. Sci., Oklahoma Univ., Norman, OK, USA

fYear

1989

fDate

14-16 Aug 1989

Firstpage

752

Abstract

The design and implementation of an automatic controller such as a PID (proportional, integral, and derivative) controller or a self-tuning regulator has been of much interest in process control, and has been the subject of research in the past decade. Many of the controllers in use today require manual tuning of the gain constants. Using the inverted pendulum as an application, it is possible to use a heuristic to both balance the pole and restore the cart to the proper position. It is shown that an artificial neural network (ANN) can be used as the heart of the controller. The ANN controller can be trained by the same heuristic to balance the broom

Keywords

adaptive control; feedback; neural nets; self-adjusting systems; three-term control; PID; artificial neural network; automatic controller; feedback controller; heuristic; inverted pendulum; neural network; process control; self-tuning regulator; Adaptive control; Artificial neural networks; Automatic control; Neural networks; PD control; Pi control; Process control; Proportional control; Three-term control; Tuning;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 1989., Proceedings of the 32nd Midwest Symposium on

Conference_Location

Champaign, IL

Type

conf

DOI

10.1109/MWSCAS.1989.101964

Filename

101964