An optimization based approach for evalutation of vehicle rollover propensity

Testing and evaluation of engineering systems are expensive and time-consuming, yet very crucial tasks during the development of engineering products. Testing of dynamic systems is performed to ensure that designs meet requirements for performance, safety, durability, and reliability. For safety-critical dynamic systems, where human life and environmental integrity are at stake, the need to prevent errors necessitates rigorous examination and the search for the worst possible scenarios.; Over the last two decades, computer-aided engineering has led to tremendous progress in many industries, which is evident in the growing popularity of computer-based simulation models to replace early production prototypes. We contribute to this line of research by proposing an approach which utilizes optimization techniques to provide a framework for a more rigorous examination procedure. Specifically, this dissertation focuses on the applications of optimization techniques in the vehicle active safety system design and testing.; In the first half of this work, an innovative evaluation procedure based on worst-case evaluation methodology is developed. Worst-case evaluation method is used to investigate the effects of vehicle and environment parameters and the influence of vehicle dynamic control (VDC) systems on the roll dynamics of the Sport Utility Vehicles. Worst-Case Analysis Toolbox (WCAT), an object-oriented worst-case scenario generation package written in Matlab, is developed to automate the testing of dynamic system characteristics.; The second half of the dissertation concentrates on the design of a new human driver model and its integration into a computer-based design and testing process. The goal of the driver model is to simulate the dynamics of a vehicle under a broad spectrum of driver steering inputs and analyze the vehicle/controller/environment interactions leading to improvement in the design process for the vehicle and active safety systems.; The worst-case analysis methodology and the driver model are then utilized to formulate an integrated computer-based design and testing process. Using an example iterative loop, the performance of a vehicle and VDC system design are evaluated. The benefits of including worst-case analysis within the design process are presented.