A new modeling method for nonlinear rate-dependent hysteresis system based on LS-SVM

In this paper, a novel modeling method based on least square support vector machines (LS-SVM) is proposed to deal with the rate-dependent hysteresis system. It is possible to construct a unique dynamic model in a given frequency range for a rate-dependent hysteresis system using the selected compound frequency as the training set of LS-SVM, which guarantees an outstanding generalization ability of frequency. The precision of the model attributes to the improved strategy of parameter regulation of kernel parameter and penalty factor, which is also address in the paper. Simulations on a Giant Magnetostrictive Actuator (GMA) verify both the effectiveness and practicality of the proposed modeling method.