Spatially Formulated Connected Automated Vehicle Trajectory Optimization with Infrastructure Assistance

This paper presents a constrained connected automated vehicles (CAVs) trajectory optimization method on curved roads with infrastructure assistance. Specifically, this systematically formulates problems in spatial domain and curvilinear coordinate. As an alternative of temporal Cartesian coordinate formulation, our formulation provides the flexibility to describe complex road geometries, traffic regulations, obstacles, which are usually spatially varying rather than varying, assistances vehicle (V2I) communication. Based we first conducted mathematical proof controllability system, show that system can be controlled Further, multiobjective model predictive control (MPC) approach is designed optimize trajectories rolling horizon fashion satisfy collision avoidances, kinematics constraints simultaneously. To verify efficiency method, multiscenario numerical simulations conducted. Suggested by results, proposed provide smooth vehicular trajectories, avoid simultaneously follow regulations different scenarios. Moreover, robust change geometries other potential disturbances curvature, work zone, speed limit change.