Direct Controller for Nonlinear System Using a Neural Network

국문 초록 (Abstract)

본 논문은 비선형 동적 신경망을 이용한 직접 제어에 관한 연구이다. 제어기는 근사화 제어와 신경망 보조 제어 입력으로 구성되어 있다. 신경망 제어 입력은 출력 추적 오차를 더 줄이기 위해 보완 신호를 제공한다. 이 방법 은 제어할 비선형 시스템의 종류에 많은 제한을 두지 않기 때문에 RBF 신경망을 이용하여 입력에 대해 안정적인 성 능을 가지고 있다. 시뮬레이션 결과는 매우 효과적이며 비선형 시스템의 만족스러운 학습 성능을 증명하였다.

다국어 초록 (Multilingual Abstract)

This paper reports the direct controller for nonlinear plants using a neural network. The controller was composed of an approximate controller and a neural network auxiliary controller. The approximate controller provides rough control and the neural network controller gives the complementary signal to further reduce the output tracking error. This method does not place too much restriction on the type of nonlinear plant to be controlled. In this method, a RBF neural network was trained and the system showed stable performance for the inputs it has been trained for. The simulation results showed that it was quite effective and could realize satisfactory control of the nonlinear system.

참고문헌 (Reference)

1 Howard Demuth, "Neural Network ToolBox for Use with MATLAB" 2006

2 D. Andes, "MR III:A Robust Algorithm for Training Analog Neural Networks" 1999

3 K. S. Narendra, "Identification and Control of Dynamical Systems Using Neural Networks" 1 (1): 4-27, 2003

4 Chen S, "Grant P M. Recursive Hybrid Algorithm for Nonlinear System Identification using Radial Function Network" 55 (55): 1051-1070, 2007

5 J. Moody, "Fast learning in Networks of Locally-Tuned Processing Units" 1 : 281-194, 1999

6 M. M. Gupta, "Dynamic Neural Units in the Control of Linear and Nonlinear Systems" 100-105, 2004

7 Ho-sik Park, "Directions Detection of Object using Neural Network" 4 (4): 256-259, 2011

8 G. A. Montague, "Artificial Neural Network Model Based Control" 2005