영상 센서와 가상 차선을 이용한 자동 차선 변경 알고리즘에 대한 연구

Abstract

Advanced driver assistance systems(ADAS) (e.g., adaptive cruise control(ACC), lane keeping assist system(LKAS), etc.) have been developed for safety and convenience, and actively utilized for the development of an autonomous driving system. In particular, automated lane change is a key technology of the lateral control area in an autonomous driving system. In this study, automated lane change algorithm using a vision sensor and a pseudo-lane is suggested in order to generate a path for lane change. The path planning for lane change is proceeded in two phases. In phase 1, a path is generated with line information at each time step. When the host vehicle enters the transition area between the two lanes, the line information sometimes has a low reliability. Therefore, the pseudo-lane path based on vehicle motion is generated in order to solve the problem in phase 2. The proposed algorithm is first validated in simulations and the experimental verification is carried out.