01697384

Component

This research proposed an optimal control based motion planning algorithm for Connected and Automated Vehicles (CAVs) free cruising on a highway with partially connected and automated traffic. It overcomes the shortcomings of conventional methods and is able to: i) fulfill certain objective according to users’ preference, such as transportation mobility, fuel efficiency and safety; ii) function under partially connected and automated environment; iii) consider environment stochasticity into optimization; iv) automate both longitudinally and laterally; v) be computational efficient for potential field implementation; vi) guarantee optimal solution. A new problem formulation framework named Model Predictive Control with Stochastic Constraint (MPCSC) was proposed. It was designed to consider stochasticity of adjacent human driven vehicles by including probability constraints into optimization. The proposed MPCSC was solved using a Chang-Hu’s algorithm which was previously developed by this research team. MPCSC was evaluated against conventional Model Predictive Control (MPC). The results confirm that the proposed MPCSC outperforms conventional MPC. It is collision-free while MPC shows 65% more collisions, 75% more direction changes and 15.7% smaller standard deviation of longitudinal speed. The average computation time of MPCSC is approximately 0.1 seconds when running on a laptop equipped with an Intel i5-3230M CPU.

This paper was sponsored by TRB committee AHB30 Standing Committee on Vehicle-Highway Automation.

19-03877

English

0000-0002-0144-9489

5p

2019

Transportation Research Board 98th Annual Meeting

Digital/other

Figures; References

Advanced vehicle control systems; Algorithms; Connected vehicles; Highway traffic; Intelligent vehicles; Stochastic processes; Vehicle mix

Model predictive control

Data and Information Technology; Highways; Vehicles and Equipment

Transportation Research Board Annual Meeting 2019 Paper #19-03877

TRIS, TRB, ATRI

Dec 7 2018 9:26AM