Fault Identification for High-Speed Vehicle Suspension System Using Nonlinear Filtering

Suspension system is vital important to safe operation of the high-speed vehicle. Parameter estimation is a new way for fault identification of suspension system by monitoring parameter attenuation and sudden failure of key components. A lateral dynamic model is built in this paper, based on which a nonlinear filtering algorithm called Rao-Blackwellized Particle Filter (RBPF) is applied for parameter estimation. Furthermore, RBPF based on repeat-uniform-sampling strategy is proposed for avoiding impoverishment of parameter particles, and fault identification of sudden failure is realized. Simulation results show that algorithms proposed and studied in this paper are effective and reliable with high accuracy.