Motion prediction of a heavy vehicle to avoid understeer and rollover: Modelling and simulations of a heavy rigid truck with motion prediction and control envelope to maintain safe driving conditions

Abstract

This thesis concerns an investigation of a safety system to avoid rollover and understeering in a heavy vehicle. This safety system includes a motion prediction with implemented braking action, as well as a yaw controller. The motion prediction has a 2 second prediction horizon, and the braking action is implemented in all wheels. The motion prediction includes a model of the vehicle based on Newtonian formalism and the tire model is based on the Pacejka 2002 model. The yaw controller is a QFT design, and is implemented on the rear wheels. The contribution to the field is the investigation of a real-time motion prediction safety system, which aims to reduce the risk of rollover and understeering under normal driving conditions. The objective is that this system would predict and implement a preventive action sooner than a driver realistically could, thus reducing the risk of a critical traffic situation. The thesis work has been conducted through computations and simulations in Matlab and Simulink, and simulations in C++.