Model-based precision position and force control of SMA actuators with a clamping application

This paper presents a novel control strategy for precision control of Shape Memory Alloy (SMA) actuators by using a combination of large and small-signal controllers. The large-signal controller, responsible for coarse positioning, is a nonlinear PID (NPID) temperature controller that gets its desired temperature value online from a computationally efficient stress-strain and phase kinetics SMA model. The small-signal controller, responsible for fine positioning, is variable structure controller that uses an asymmetrical boundary layer to prevent over-heating of the wire. Further, the control signals adapt to the dynamic load on the wire to prevent overshoots. The control results show precise tracking control with exceptional disturbance rejection for various loads. The control strategy is then used for force control in an SMA-based clamping vice to clamp a thin-walled glass test tube. While the large-signal controller remains unchanged, the small-signal controller is used as a force controller rather than a position controller. The control unit of the vice is an embedded processor, on which the control algorithms run with a sample time of 3 ms. The vice is used to clamp the test tube at various desired step and sinusoidal forces and shows accurate force tracking and good disturbance rejection.