A recurrent neural net approach to one-step ahead control problems

Author

Yip, Percy P C ; Yoh-Han Pao

Author_Institution

Dept. of Electr. Eng. & Appl. Phys., Case Western Reserve Univ., Cleveland, OH, USA

Volume

24

Issue

4

fYear

1994

fDate

4/1/1994 12:00:00 AM

Firstpage

678

Lastpage

683

Abstract

In this paper, we present a recurrent neural net technique to provide control actions for nonlinear dynamic systems. In most current neural net control approaches, two nets are usually required. One acts as a system emulator, and the other one is a controller network. Rather than using two nets, our system requires only one net which is the system emulator. In our proposed system, a neural net is used to learn the forward dynamics of the system, and the control signal is evolved from the output of the same net with use of an equation of motion. There is no need to learn the control law from another neural net, such as a system inverse net. The use of the proposed algorithm is illustrated with an example

Keywords

learning (artificial intelligence); nonlinear dynamical systems; nonlinear systems; recurrent neural nets; equation of motion; forward dynamics learning; nonlinear dynamic systems; one-step ahead control problems; recurrent neural net; system emulator; Adaptive control; Control systems; Equations; Error correction; Motion control; Neural networks; Nonlinear control systems; Nonlinear dynamical systems; Physics; Recurrent neural networks;

fLanguage

English

Journal_Title

Systems, Man and Cybernetics, IEEE Transactions on

Publisher

ieee

ISSN

0018-9472

Type

jour

DOI

10.1109/21.286388

Filename

286388