Prediction of magnitude data of multi-channel vibration on a vehicle platform environment

The vibration prediction model of the key vibration source points of a vehicle platform is studied. Starting from the analysis of the factors affecting the vibration response and the correlation analysis, the key factors affecting the vibration response of the same vehicle structure are sorted out. The vibration prediction model of the key vibration source points is established by using neural network, and using the measured data to verify the accuracy of the vibration prediction model of the cabin section I, cabin II and cabin III of the platform. The results show that in the frequency range of 1∼500Hz, the maximum error of root mean square of cabin I is 2.75dB, that of cabin II is 2.3dB, and that of cabin II is 2.97dB.