Application of Laguerre based adaptive predictive control to Shape Memory Alloy (SMA) Actuator

This paper discusses the use of an existing adaptive predictive controller to control some Shape Memory Alloy (SMA) linear actuators. The model consists in a truncated linear combination of Laguerre filters identified online. The controller stability is studied in details. It is proven that the tracking error is asymptotically stable under some conditions on the modelling error. Moreover, the tracking error converge toward zero for step references, even if the identified model is inaccurate. Experimental results obtained on two different kind of actuator validate the proposed control. They also show that it is robust with regard to input constraints.