Autonomous Electric Vehicle Route Optimization Considering Regenerative Braking Dynamic Low-Speed Boundary

Abstract

Finding the optimal speed profile of an autonomous electric vehicle (AEV) for a given route (eco-driving) can lead to a reduction in energy consumption. This energy reduction is even more noticeable when the regenerative braking (RB) capability of AEVs is carefully considered in obtaining the speed profile. In this paper, a new approach for calculating the optimum eco-driving profile of an AEV is formulated using mixed-integer linear programming (MILP) while carefully integrating the RB capability and its limitations in the process of obtaining a driving profile with minimum energy consumption. One of the most important limitations of RB which has been neglected in previous studies is operation below the low-speed boundary (LSB) of electric motors, which impairs the energy extraction capability of RB. The novelty of this work is finding the optimal speed profile given this limitation, leading to a much more realistic eco-driving profile. Python is used to code the MILP problem, and CPLEX is employed as the solver. To verify the results, the eco-driving problem is applied to two scenarios to show the significance of considering a dynamic LSB. It is shown that for the route under study, up to 27% more energy can be harvested by employing the proposed approach.

Keywords

• autonomous electric vehicle
• eco-driving
• regenerative braking
• low-speed boundary
• mixed-integer linear programming