A Methodology for Threat Assessment in Cut-in Vehicle Scenarios 2021-01-0873

Advanced Driver Assistance System (ADAS) has become a common standard feature assisting greater safety and fuel efficiency in the latest automobiles. Yet some ADAS systems fail to improve driving comfort for vehicle occupants who expect human-like driving. One of the more difficult situations in ADAS-assisted driving involves instances with cut-in vehicles. In vehicle control, determining the moment at which the system recognizes a cut-in vehicle as an active target is a challenging task. A well-designed comprehensive threat assessment developed for cut-in vehicle driving scenarios should eliminate abrupt and excessive deceleration of the vehicle and produce a smooth and safe driving experience. This paper proposes a novel methodology for threat assessment for driving instances involving a cut-in vehicle. The methodology takes into consideration kinematics, vehicle dynamics, vehicle stability, road condition, and driving comfort. General kinematics have been used to define the longitudinal threat assessment level and the performance of external disturbance attenuation in robust control theory has been used to define the lateral threat assessment level. Subsequently the threat assessment level has been analyzed in various cut-in vehicle scenarios. The simulation results validate the effectiveness of the proposed method for threat assessment in response to cut-in vehicles. The proposed threat assessment method will help further develop safe and smooth ADAS reactions to cut-in vehicles.