Statistics-based simulation methodology for evaluating collision countermeasure systems performance

Extensive dynamic simulation studies were conducted to evaluate the ability of collision countermeasure systems (CMS) to prevent run-off-road accidents. A key challenge facing the development of an effective CMS is the ability to function successfully given the wide range of driver types, vehicle characteristics, road conditions and driving situations. A Monte Carlo approach based on Latin hypercube sampling was implemented so that influences on CMS effectiveness could be evaluated efficiently. The results demonstrate the effectiveness with which the methodology can be used to assess CMS performance, and indicate the need for an adaptive CMS that anticipates and responds correctly to diverse and rapidly-changing driving scenarios.