Statistical modelling of predicted non-stationary vehicle vibrations

This paper describes a method to predict the vertical vibrations of road vehicles from measured pavement profiles. It discusses the limitations of current methods used for analysing and simulating vehicle vibrations and shows that more accurate characterization and simulation of the transport environment must take into account the non-stationary nature of road vehicle vibrations. Vertical vibrations for typical transport vehicles under various operating conditions and pavement profiles are predicted using a computer model of the vehicle characteristics and analysed to produce the spectral and statistical characteristics. The paper also presents an improved method to compute the vibration intensity by using a dynamic segmentation data reduction technique. The effectiveness of the procedure to characterize the non-stationarity of random vehicle vibrations is demonstrated. Finally, the paper deals with the statistical distribution of the vibration intensity and demonstrates how it can be adapted to a technique for the simulation the non-stationary nature of random vehicle vibrations