Research on giant magnetostrictive actuator online nonlinear modeling based on data driven principle with grating sensing technique

Abstract

A novel Giant Magnetostrictive Actuator (GMA) experimental system with Fiber Bragg Grating (FBG) sensing technique and its modeling method based on data driven principle are proposed. The FBG sensors are adopted to gather the multi-physics fields' status data of GMA considering the strong nonlinearity of the Giant Magnetostrictive Material and GMA micro-actuated structure. The feedback features are obtained from the raw dynamic status data, which are preprocessed by data fill and abnormal value detection algorithms. Correspondingly the Least Squares Support Vector Machine method is utilized to realize GMA online nonlinear modeling with data driven principle. The model performance and its relative algorithms are experimentally evaluated. The model can regularly run in the frequency range from 10 to 1000 Hz and temperature range from 20 to 100 °C with the minimum prediction error stable in the range from −1.2% to 1.1%.