Planning The Velocity of a Parallel Hybrid Electric in Vehicle-to-vehicle Autonomous Driving: an Optimization-based Approach

Improved numerical tools are required to foster flexible and effective advancement of innovative electrified and highly automated road vehicles. This paper proposes an optimization-based approach to off-line plan the longitudinal velocity of a hybrid electric vehicle (HEV) when travelling as Ego vehicle in a vehicle-to-vehicle (V2V) autonomous driving scenario. A parallel P2 hybrid powertrain layout is retained along with the corresponding on-board supervisory controller. A mathematical formulation for the optimal V2V autonomous driving control problem is provided and consequently solved with an optimization method based on dynamic programming (DP). The implemented DP formulation particularly exploits information about the overall longitudinal speed profile of a Lead vehicle in a predefined driving mission to determine the velocity profile of the Ego vehicle. Optimization constraints involve maintaining the inter-vehicular distance value within allowed limits while aiming at minimizing both the magnitude of Ego vehicle acceleration events and the overall Ego vehicle fuel consumption as predicted according to the on-board hybrid supervisory control logic. Simulation results for different driving missions demonstrate that, using the proposed DP formulation, the Ego vehicle can achieve both smoother speed profiles and improved fuel economy by some percentage points in V2V autonomous driving compared to the retained Lead vehicle embedding the same HEV powertrain layout.