Model Reference Adaptive Sliding Surface Design for Nonlinear Systems

The paper introduces a new model reference adaptive sliding surface design algorithm in order to determine possible nonlinear time-varying “sliding surface” for a general class of nonlinear systems. A model reference sliding surface is first designed for a reference nonlinear system by using the state-dependent Riccati equation techniques. Then, the sliding surfaces for nonlinear systems are obtained from the reference sliding surface by using an adaptation rule. The adaptation rule to satisfy the convergence of sliding surface for the nonlinear system to that of reference one is derived. The sliding mode controller for the nonlinear plant is then designed by using the adaptive sliding surfaces. The proposed method is illustrated with an autopilot design for different missile models.