In order to overcome the interference of environmental noise and improve the quality of EMD, the wavelet method is used to eliminate the noise signal. The correlation coefficients are applied to remove the pseudo components. And then, the random decrement method and ITD method are used to identify the natural frequency and damping ratio of train’s floor. Compared the identification results with those from LMS commercial software, it is found that the relative error of frequency is less than 5%, the relative error of damping ratio is less than 7%. This shows that the method has high precision of identification.

The structure of High-speed vehicle is complex, though simulation to obtain structural properties on the current computer simulation is still difficult.Modal testing is used to obtain dynamic characteristics of the structure. This paper describes a test method for structural characteristics, and its realization of the basic algorithm. Multi-excitation method is to be used and the measuring point for the minimum non-resonant component of the objective function of the adjustment force method is to establish. Simulation results show this method is simple, less post-processing work,and high reliability.