The acceleration and force sensors are taken as the measurement parameter,vibration power flow analysis is conducted to the data collected using MATLAB software,the results show that validity of engine suspension vibration is intuitively judged using the power flow,it is a evaluation index,moreover,it is not only convenient measurement,but also the influence of phase difference between the measured parameters to vibration isolation effect is not needed consider.

The damping effect of 485 diesel engine is taken as research object, on the same conditions, bench test is conducted respectively to self-developed MR fluid damper and traditional rubber damper, comparative analysis of domain and spectrum are conducted to collected test data using MATLAB, the results show that damping effect of MR fluid damper to engine is significantly superior to effect of traditional rubber damper.

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.