A Computationally Efficient and Hierarchical Control Strategy for Velocity Optimization of On-Road Vehicles

Lulu Guo,Hong Chen,Qifang Liu,Bingzhao Gao

DOI: https://doi.org/10.1109/tsmc.2018.2826005

2018-01-01

IEEE Transactions on Systems Man and Cybernetics Systems

Abstract:Velocity profile optimization of on-road vehicles is one of the main eco-driving techniques, which has great potential to extend the capability of powertrain and automatic longitudinal control by minimizing the energy consumption. Due to the multi factors affecting the driving trajectory and longer prediction horizon comparing with other traditional control, the calculation of a velocity profile optimization often requires a large number of computations. In this paper, a hierarchical control (HC) strategy of velocity optimization is proposed to reduce computation burden with little accuracy loss. In the HC strategy, a specific driving task is divided into several operation of modes as acceleration (A), constant speed (C), deceleration (D), and braking (B). The shift timing of the driving modes are optimized by formulating a nonlinear programming problem in a master controller. Then, engine torque, gear position, and brake force are optimized in each driving mode. Results indicate that the computation time of velocity profile optimization using the proposed HC strategy is reduced by 90% of the ones using the basic centralized optimal controller while the resulting velocities are similar. It is also shown that an improvement of 30% in fuel economy is achieved compared with the real-life human-driven velocity profiles.