Linearization of a current-driven reluctance actuator with hysteresis compensation

This paper investigates the influence of hysteresis present in the ferromagnetic core of a variable reluctance actuators on the force reproducibility. To reduce this influence and to boost reproducibility, a hysteretic inverse actuator model is derived and used to linearize a current-driven reluctance actuator. Furthermore, an identification procedure for identifying the parameters of the hysteresis model and the remaining actuator non-linearities is presented. Two actuators are experimentally tested with the proposed compensator and a linearization error smaller than 0.05% of the maximum force is achieved, which is an order of magnitude improvement over single-valued inverse compensators. A comparably small error is obtained for non-trivial, non-periodic inputs when higher order reversal curves of the actuator hysteresis have to be reproduced as well. The simple structure of the compensator allows a fast implementation in digital controllers.

► Hysteresis in current-driven E-core and C-core reluctance actuators is studied and modeled.
► Actuator models which neglect hysteresis have limited predictability.
► Predictability can be improved by using the parametric hysteresis inverse.
► The linearization error due to hysteresis is reduced by an order of magnitude.
► The proposed inverse model is intuitive, and easy to identify and implement.