Abstract: A polynomial dimensional decomposition pseudo-excitation method (PDD-PEM) was established in the frequency domain to quantify the uncertainty of random vibration power spectrum density for vehicles with uncertain parameters under random road excitation. Transforming stationary random vibration analysis into harmonic load analysis through pseudo-excitation method, and transforming double random problems into single random problems; At the same time, the polynomial dimensional decomposition method is used to construct a random Surrogate model, and the explicit function of the power spectrum density response expressed by the polynomial basis is given, which effectively realizes the probabilistic evaluation of the structural response in the uncertain parameter space. In numerical examples, the method established in this paper was used to analyze the random vibration of vehicle systems with uncertain parameters under road roughness. Compared with the Monte Carlo method, the correctness and effectiveness of the established method were verified, and the influence of uncertain parameters on the structural response statistical characteristics was further discussed. These works laid a certain foundation for considering the optimization and control problems of vehicle system parameters with uncertainty.