가상예측차선을 적용한 멀티레이트 예측기 개발

Abstract

In autonomous driving vehicle control, the vehicle uses a vision sensor to get a road information and inertial measurement unit to get a vehicle motion information. In general, speed of vision sensor update is slower than motion sensor and non-uniform. Thus the system needs method making uniformly update data and sampling period of estimator. We propose predictive virtual lane and the multirate estimator to solve the problems. The performance of the proposed control methods is validated via simulations implemented in CarSim and Matlab/Simulink.