High-Precision Flux Linearization of Linear Reluctance Actuators With Short Stroke and High Force Density

The linear reluctance actuator, renowned for its exceptional output density, emerges as the most promising option for high acceleration demands in next-generation lithography systems. However, the inherent hysteresis nonlinearity between current and magnetic flux density significantly impacts precision force output and restricts its application in nanometer positioning. Specifically, this limitation becomes particularly pronounced when hysteresis is coupled with eddy effects and motion displacement, resulting in a rate- and displacement-dependent hysteresis effect within the actuator. To address this complex hysteresis nonlinearity, we consider the model alteration induced by displacement as a disturbance effect of the rate-dependent system. Subsequently, we propose a compound control method that integrates iterative feedforward tuning and inverse model feedback. Unlike existing control methods heavily reliant on accurate system models for performance optimization, our proposed strategy achieves: 1) reduced dependence on system model information through data-driven estimation; 2) improved adaptability to different trajectories through feedforward parameterization; and 3) enhanced robustness through iterative learning. Experimental results not only demonstrate the effectiveness of our proposed control method in mitigating hysteresis but also validate its suitability for various trajectories.