Modeling and Identification of MSMA Self-sensing Actuator Based on Hammerstein Model

A magnetically controlled shape memory alloy (MSMA) is an innovative smart material characterized by significant output strain, quick response, and high energy density. Under environmental temperatures, its shape changes can be controlled by putting in for a magnetic field or external stress. It is considered to be an ideal intelligent material for self-sensing actuator. The most notable feature of MSMA is its memory capability, which allows it to retain its shape after the external magnetic field and mechanical stress are removed. Additionally, it exhibits reversible characteristics, enabling MSMA to produce deformation output displacements and forces generated under the influence of a magnetic field. The input and output of the MSMA self-sensing actuator exhibit a complex nonlinear relationship. To analyze these nonlinear characteristics, a multivariable Hammerstein model is developed, with its unknown parameters identified using the sand cat swarm optimization (SCSO) algorithm. To address the issues of low precision in parameter estimation and dull convergence in the chaotic sand cat swarm optimization (SCSO) algorithm, an enhanced chaotic sand cat swarm optimization (CSCSO) algorithm is proposed. The cubic chaotic map is incorporated into the identification process to increase the diversity of the initial sand cat population and enhance the algorithm's capability to avoid local optima. The algorithm can improve its capability to evade local optimal solutions. Finally, the model calculation results are compared against experimental data to validate both the model's correctness and the improved algorithm's accuracy in model identification, effectively addressing the issue of model precision.