Automated vehicle steering control has been actively researched in the automotive industries and academia over a decade. While several automotive companies and suppliers have recently demonstrated autonomous parking, lane keeping control, and lane centering control systems, automated lane change and obstacle avoidance maneuvering have not been as well demonstrated with the same level of maturity. This paper describes an algorithm that assesses environment and situation around the subject vehicle and makes a proper decision when an automated lane change or obstacle avoidance maneuvering is needed. The algorithm continuously monitors the surrounding traffic and lane markings using various types of sensors, and makes judgments along the vehicle future motion. Collision threat is evaluated by comparing the future path of the vehicle and the surrounding traffics in temporal-spatial plane. Typical driving behavior patterns are modeled to ensure safety under various scenarios. This algorithm has been implemented on a test vehicle and validated on straight and curved roads for various speeds of up to 110km/h. Several test cases have been completed and the results are provided.

This content is only available via PDF.
You do not currently have access to this content.