Mixed-kinematic body sideslip angle estimator for high performance cars

Body sideslip angle estimation is a well-known issue in the automotive field. Many vehicle dynamic control systems are based on the control or monitoring of the sideslip angle. Unfortunately, measurement of this quantity requires expensive instrumentation that commercial cars can not be equipped with. This paper further develops the so-called hybrid dynamic-kinematic approach to the problem providing an automated tuning approach. A state observer merges a kinematic model with a lateral dynamic one, so that they can compensate each other shortcomings. The resulting estimator depends on a number of parameters whose manual tuning require expertise. The paper proposes a grey-box approach for the tuning of the algorithm. The optimization-based approach requires only standard track tests and no special vehicle characterization. The approach is validated on a large experimental dataset from a high performance car.