Parameters identification of rational function power spectral density of pavement based on genetic algorithms

Aiming at the parameter identification problem of power spectral density (PSD) of pavement second-order rational function, an optimization calculation model was established, in which the minimum mean-square-error of second-order rational function PSD and power function PSD given by ISO was chosen as the optimization objective. The optimum parameters of A- to D-grade pavement second-order rational function PSD were obtained by the genetic algorithms. On this basis, a road roughness model was established based on the basic principles of harmonic superposition. Then the spectrum analysis of the established road roughness model was made using auto regressive (AR) model power spectrum density estimation method. Results indicated that the pavement second-order rational function PSD was well coincidental with standard road classification power function PSD given by ISO, and the parameters of pavement second-order rational function PSD could be accurately identified using the genetic algorithms.