Performance-based model-free adaptive finite-time control for shape memory alloy actuated swing platform

Abstract

Shape memory alloy (SMA), a kind of smart material, can be used as an actuator in many fields; however, its strong nonlinearity and parameter uncertainty hinders its application in high-tracking accuracy tasks. This paper addresses the tracking control problem of the SMA actuated swing platform suffering from completely unknown nonlinear model information and prescribed finite-time error constraints. First, the equivalent dynamic linearization model of the swing platform is established, and the unknown disturbance is estimated and compensated by the extended state observer. Meanwhile, a novel discrete-time performance function is proposed, and the prescribed finite-time tracking error constraints are transformed into a new equivalent unconstrained task. Second, the model-free adaptive sliding mode controller is designed using the unconstrained error, and the complete control law, including sliding mode control law and equivalent control law, is derived. Third, the stability of the closed-up swing platform is guaranteed by employing the Lyapunov method. Finally, experiments reveal that the proposed method is preferable.