Adaptive Neural Network Controller Design for a Class of Nonlinear Systems Using SPSA Algorithm
Ching Hung Lee, Tsung Min Yu, Jen Chieh Chien · 2014
Abstract—In this paper, we propose a novel SPSA-based on-line adaptive decoupled control scheme by using PID neural network for a class of nonlinear systems. In addition, the update laws of parameters with adaptive optimal learning rate are proposed based on the Lyapunov stability theorem, this guarantees the stability of closed-loop system. In addition, the affect of the frictional force model and uncertainty are discussed and analyzes. The proposed approach is applied in the translational oscillations with a rotational actuator (TORA) system. In experimental results, the proposed control is realized by DSP to demonstrate the performance and the efficiency. Index Terms—adaptive control, PID neural network, simultaneous perturbation stochastic approximation, real-time control. R I.