Trajectory Planning for Connected and Autonomous Vehicles at Freeway Work Zones under Mixed Traffic Environment

Abstract

Mixed traffic control with connected and autonomous vehicles (CAVs) and human driving vehicles (HVs) is becoming a hot topic. CAV trajectory planning at work zones under the mixed traffic environment is a big challenge. Existing studies only focus on longitudinal trajectories (e.g., acceleration profiles), ignoring lateral trajectories (lane changing). This study proposes a trajectory planning model for CAVs at work zones under mixed traffic environment. Both longitudinal and lateral trajectories are considered. On the basis of the states of CAVs and of HVs observed by CAVs, the number and initial states of unobservable HVs in the planning horizon are estimated considering the interactions between vehicle driving behaviors. A trajectory planning model is then formulated to optimize acceleration profiles and lane choices of CAVs in the planning horizon in a centralized way. The minimization of total vehicle delay and fuel consumption is adopted as the objective function. A car-following model and a lane-changing model are adopted to capture the driving behaviors of HVs. The proposed model is a mixed-integer linear program. Numerical studies validate the advantages of the proposed trajectory planning model over late merge control for vehicle delay and fuel consumption.