Identification of physical parameters of a passenger car

This paper deals with the identification of the physical parameters of a passenger car. After studying the local structural identifiability of two models of different complexity, the identifiable parameters which are unknown or difficult to measure are identified by minimising a multivariable, cost criterion in terms of maximum likelihood. The choice of the optimisation algorithm is considered, before realistic, efficient quantification of the uncertainties in the estimation of parameters is proposed. This approach is applied to a road vehicle. Results in the form of predictive capability and precision for the estimated parameters, illustrate the precision and robustness of the model identified in this way.