Region Tracking Control for Nonlinear Systems Based on Stochastic Resonance

This paper investigates region tracking problem for a nonlinear system subject to measurement noise. Its purpose is to make enough use of the prescribed boundaries to reduce the fluctuations of control inputs and energy consumption. A novel region tracking control scheme is proposed based on bistable stochastic resonance. To construct a bistable stochastic resonance model, the tracking error is treated the Brownian particle and a weak periodic signal is added to the model, while the error in the dynamic loop is considered as noise. If the potential parameters are properly selected, the optimal stochastic resonance of the tracking error occurs. Then, the tracking error will switch between the potential wells determined by the potential parameters. Finally, the new design is applied on an over-actuated underwater vehicle. The results verify the superiority of the proposed controller in region tracking performances, including the use of the prescribed boundaries and fluctuations of control inputs.