A Bayesian framework-based vehicle parameters identification method with unknown road excitation

Accurate information on vehicle parameters is essential for vehicle design, vehicle-bridge interaction analysis, and road maintenance. Currently, most vehicle parameters identification (VPI) methods are deterministic and rely on strict experimental conditions. This paper proposes a simple method based on Bayesian framework to identify uncertain vehicle parameters, which only requires the vibration response of the vehicle under arbitrary excitation. First, the dynamics of the vehicle is analyzed and the link between excitation and response is established using the vehicle frequency response function. Subsequently, the likelihood function is formulated based on the difference function between the measured responses and the expected responses of updated model. Notably, this method introduces a creative probabilistic form of the frequency-domain response assurance criterion. Furthermore, a numerical simulation of the vehicle driving over bumps is performed to assess the influence factors on VPI, such as noise pollution and the excitation correlation between the front and rear wheel. At last, the parameters of a van are identified through field test and the reliability of the results is demonstrated.