Output fluctuation and overshoot restraining model-free adaptive control for a class of discrete-time nonlinear single-input single-output systems

This paper proposes a novel model-free adaptive control approach to restrain system output fluctuation and overshoot (SOFO) for a class of discrete-time nonlinear single-input single-output systems. First, historical input/output sensor data is utilized to online construct an equivalent explicit data model for control design, making the controller model-free. Second, a novel compensation input term including a compensation gain and a signum function is proposed to soften system input, which in turn restrains the SOFO. The signum function is parameterized by incremental output. The controller and adaptive law of the compensation gain are designed according to the discrete-time Lyapunov stability theories. Last, numerical simulations validate that the proposed approach can restrain the SOFO successfully and reduce response time compared with the existing approaches.