Rule-Based Method for Collision Warning and Active Control of Automotive Safety Systems

摘要

Collision countermeasure and warning systems are under development in the automotive industry for enhanced occupant safety. Sensor technologies incorporating radar, and vision systems, obtain information about potential obstacles in the external environment of the host vehicle. Safety critical decisions are made based on detected obstacles, and countermeasures are systematically activated. Important performance specifications of an adaptive safety system include its ability to tailor deployment decisions on occupant and collision-type characteristics, and decision-making robustness. This paper presents a rule-based methodology for automotive collision warning and brake-assist systems for potential improvement of occupant protection systems. The system primarily focuses on assessing the potential for a collision to occur, and robust decision-making to warn, and then actively adapt safety systems to potential collision conditions. Key requirements for the adaptive safety system design are discussed. Results from application of the rule-based system are presented.