Spatial Domain Reconstruction of Road Roughness Based on White Noises Filtering

Based on lower and upper triangular matrix (LU) decomposition of the power spectral density (PSD) matrix of road roughness for different wheel tracks, the transfer functions of road roughness was presented for different wheel tracks to two independent white noises signals. The coherent function in transfer functions was rationally approximated by Pade expansion. In addition, the complex transcendental function related to the coherent function in the transfer functions was rationally approximated by Chebyshev-Pade expansion. Based on white noises filtering (WNF) method, a 2D random road simulation model considering coherence function of different wheel tracks was constructed. The simulation results show that the Auto-PSD and the coherence function of different wheel tracks are in great agreement with the standard functions respectively. It is verified that the transfer functions derived from LU decomposition of PSD matrix of different wheel tracks and its Pade expansion and Chebyshev-Pade expansion possess high reliability in the simulation of 2D road roughness.