Topics in Modeling, Control, and Implementation in Automotive Systems

This thesis treats different aspects of automotive control-systems and consists of four papers covering different areas in this large field. The first paper presents a 9-degrees-of-freedom dynamic model of a tractorsemitrailer vehicle, aimed at simulation and evaluation of active chassis-systems for stability enhancements. Special focus is put on inclusion of the dynamics of load transfer, which are of importance in active yaw-control and roll-over prevention. The second paper presents a new easy-to-use, semi-empirical tire model for combined-slip conditions. The model is based on combining the standard rigidcarcass brush model with available empirical pure-slip models. A new method is presented for partitioning empirical pure-slip forces into components of sliding and adhesive forces. The third paper regards control of gasoline direct injection engines. Strategies based on feedback of the effective torque are shown to relax the requirements of accurate engine maps, in simulations. It is exemplified that extremum control may be used for online optimization of the engine operation with respect to fuel consumption. A new control structure that combines the extremum controller with disturbance rejection is used, to render the probing signals of the optimizing controller invisible in the engine output. The fourth paper presents a new method to limit the effects of transient faults, occurring in hardware that hosts implementations of feedback controllers. The idea is to introduce artificial signal limits that are based on closed-loop signal bounds, and combine them with an anti-windup scheme.