Vehicle dynamics modelling and validation for online applications and controller synthesis

For online automotive applications, such as vehicle dynamics control systems or state estimation, the choice of suitable mathematical models is important. In addition to model validity requirements, they have to fulfil requirements like limited computation time or numerical robustness. In this work two online suitable vehicle dynamics models with different modelling depths are discussed. First, a lumped mass model with a simplified suspension modelling technique and a complex tyre model is presented. A flatness based feed-forward and a simple feedback controller are used in order to control the vehicle speed. Additionally, a sensitivity analysis of the suspension modelling parts of the vehicle dynamics is presented. Besides the lumped mass model, a simplified model with the focus on lateral dynamics applications with a simple tyre model is introduced. The numerical properties of both models are analysed with respect to their time parameters at different speeds. An Audi A7 test vehicle equipped with precise measuring units is used in order to validate the models in the time and frequency domain. The lumped mass model shows superior validation results in lateral steady state and dynamical manoeuvres up to 2.5  Hz. The reduced model is also precise for up to 1  Hz and for steady state manoeuvres, which is sufficient for many lateral dynamics applications.