Control of an electrostrictive actuator using Newton's method

Electrostrictive actuators are a class of smart transducers with a great potential for many submicron motion applications. A major challenge for the electrostrictive actuators exists in the control of such ultra-precision motions, which are often seriously influenced by the intrinsic behaviors of electrostrictive material like non-linearity, hysteresis and creep. Based on Newton's method, this paper presents a new iterative control algorithm to improve the positioning and tracking performances of a linear multilayer electrostrictive actuator. In this algorithm, the iterative gain is not fixed but variable according to the previous output feedback and the nominal input/output relationship of the electrostrictive actuator. The convergence of this algorithm is theoretically proved quadratic and experimentally verified correct. A comparison of effectiveness of the new algorithm with that of the conventional proportional integral (PI) control and gain-fixed iterative control algorithms is made. The results show that using this new iterative control algorithm both the stability and the speed of submicron motion control have been obviously improved for the tested electrostrictive actuator.