Recent advances of information technology in automobile systems leads to the possibility of driver assistance system parameter adaptation to each individual driver and driving circumstances. This paper addresses a method of forward collision risk assessment based on real-world naturalistic driving database for advanced braking assistance system design with an individual adaptation algorithm. The risk assessment is used for driving characteristics diagnosis system, which leads to the design of headway control assistance and braking assistance systems. The algorithm is proposed as a comparison of current and normal driving behavior, which is extracted from long-term running data from the experiments. In car-following state, the forward collision threat is used as an index of hazardous potential. Finally,the discrimination of the hurry driving and the usual driving in real-world urban road driving is conducted by test drives using an experimental vehicle.
Abstract