A single auto-tuning neural controller for nonlinear dynamic systems

In this paper, a direct adaptive auto-tuning neural controller for a class of nonlinear dynamic systems is developed. Unlike traditional multi-layered neural controller, the structure of the proposed controller is simpler. The reaction time of the controller can thus be greatly reduced. Only a single neuron with three adjustable parameters is utilized in our proposed neural controller. These parameters are adjustable based on the instantaneous system output error. The adaptation law is proposed based on the Lyapunov approach. Such a controller design is very easy for hardware implementation due to its simple structure. Finally, the tracking control of the inverted pendulum system by using the proposed auto-tuning neural controller is illustrated to demonstrate the excellent control capability.