Elman Neural Network-Based Identification of Krasnosel’skii–Pokrovskii Model for Magnetic Shape Memory Alloys Actuator

Magnetic shape memory alloys (MSMAs), which are a class of innovative functional materials, are used as the actuators to be applied widely in high-precision positioning. However, the hysteresis nonlinearity in the MSMA seriously affects the precision positioning of the MSMA-based actuator. In this paper, to study the hysteresis nonlinearity in the MSMA, the Krasnosel'skii-Pokrovskii (KP) model is employed to describe the hysteresis nonlinearity in the MSMA-based actuator, and the density function of the KP model is identified by the Elman neural network. The simulations show that the modeling error rate of the KP model using the Elman neural network is 0.81%, which is reduced by 63.5% compared with that of the KP model based on a recursive least-squares method. This result demonstrates that the KP model based on the Elman neural network can accurately describe the hysteresis nonlinearity in the MSMA-based actuator.