Test Model of Automobile Engine Magneto-Rheological Mount Based on RBF Neural Network

Abstract

The nonlinear and hysteresis characteristics showed by magneto-rheological (MR) mount make it seem very difficult to establish a precise mathematical model. Based on the testing of MR mount dynamics, RBF neural network model can train and forecast the collected data. Analysis of comparing the predicting result of the RBF neural network model with the testing result shows that the trained RBF neural network model can exactly predict the dynamics of MR mount, and it provides some new ideas to implement the better intelligent control of the engine MR mount.