Cooperative weaving for connected and automated vehicles to reduce traffic oscillation

Abstract

In weaving areas, vehicles frequently carry out conflicting lane-changing manoeuvres. The frequent lane change in this area results in rapid changes in vehicles’ speed, which in turn reduces traffic efficiency and create traffic bottlenecks at weaving areas. This research proposes a cooperative weaving motion planner for connected and automated vehicles to reduce traffic oscillation. The proposed motion planner is based on model predictive control method and solved by Chang-Hu’s method. Paper presented at the Intelligent Transportation Systems (ITSC), 2018 IEEE). The motion planner only requires longitudinally automation which is accessible for most commercialized luxury vehicles. Simulation evaluation was conducted to quantify the performance of the proposed motion planner. The results show that the proposed motion planner is able to reduce traffic oscillation by 2.7% to 28.0%. Furthermore, the computation time of the proposed planner is fewer than 20 milliseconds indicating readiness to real-time application.