Bogie is an important part of railway vehicles, which has elastic components and damping components that have important influence on the safety of the vehicle. The types of the faults, the state of the faults and the selection of the parameters in the fault diagnosis methods for the bogies of rail vehicles are studied. The consequences of vehicle failure and the actual situation are analyzed.In order to realize the selection of the input parameters of the fault diagnosis algorithm and the determination of the corresponding fault location, the P index is proposed.Using the matlab models the car body and the bogie. A fault model of a vehicle suspension system is established. At the same time,fault model of the elastic components and damping components is established.The simulation of fault model can provide the data .P is used to evaluate the degree of change of satistical parameters before and after fault. At last realize the selection of the statistical parameter for the vehicle real time director diagnosis algorithm and the determination of the largest part of the vehicle vibration change. After the simulation experiment, the index P can be used to optimize the parameter input of the fault diagnosis algorithm, and determine the location of the fault.