Finite-Iteration Learning Tracking Control of Magnetic Shape Memory Alloy Actuator Based on Neural Network

Hysteresis is the key factor affecting the positioning accuracy of the magnetic shape memory alloy-based actuator (M-BA). In this paper, we investigate the finite-iteration tracking control problem of M-BA using neural network technology. Firstly, a neural network-based iterative learning control strategy is developed for finite-iteration tracking of a discrete non-affine system, where the contraction mapping principle is employed to establish the relation between tracking error and the iteration bound, thereby determining the settling iteration. Then, incorporating the mathematical induction and the data-driven methods, a sufficient condition for system convergence is provided. Finally, experiments are conducted to validate the effectiveness of the proposed method and the correctness of the theory. This study contributes to improving the positioning accuracy of the M-BA, providing insights into its potential applications in electromagnetic drive and precision electromechanical systems.