Iterative-only learning control for soft landing of short-stroke reluctance actuators

Short-stroke reluctance actuators, such as electromechanical relays and solenoid valves, often experience strong impacts at the end of standard switching operations, leading to problems like contact bounces, mechanical wear or acoustic noise. This paper introduces a novel iterative learning controller for tracking the actuator position during switching operations and ultimately decreasing the impact velocities. The control strategy is designed based on a generalized dynamical description, allowing for its application to a broad range of actuators and models. A highlight of the proposal is the absence of a real-time feedback controller, which eliminates the need for real-time measurements. Another important feature is the adaptation of the learning control gain based on a soft-landing performance index. The control performance is analyzed and validated in conjunction with an offline position estimator through Monte Carlo simulations and experimental testing. The results show that the position tracking remains accurate even in the presence of non-repeating perturbations and estimation errors, leading to a substantial reduction in switching impacts.