Vehicle tyre and handling model identification using an extended Kalman filter

This paper uses an Extended Kalman filter in an unusual way to identify a vehicle handling model and its associated tyre model. The method can be applied as an off-line batch process, or in real-time; here we concentrate on batch analysis of data from a Jaguar XJ test vehicle. The Identifying Extended Kalman Filter (IEKF) uses the full state measurement that is available from combination GPS / inertia instrumentation packs. Previous IEKF studies have shown success in identifying a bicycle model with a tyre force function for each axle. This paper extends to identification of a single, load dependent tyre model which applies to all four wheelstations, identified within a yaw-roll-sideslip model structure. The resulting model provides impressive open-loop state replication, including accurate tyre slip prediction across the fully nonlinear slip range of the tyre.