Impact of drivetrain regenerative braking assumptions on eco-driving optimal control with driver preferences

Energy efficiency in electric vehicles can be improved through eco-driving optimal control strategies. To enhance driver acceptance, recent work has incorporated driver preference models, leading to naturalistic eco-driving optimisation. However, this framework has primarily been developed at an algorithmic level and has not been evaluated using physically realistic electric vehicle models. Moreover, many existing approaches assume simplified regenerative braking behaviour, without examining how such assumptions affect optimisation performance when applied to different drivetrain configurations. To address this gap, this paper presents the first evaluation of driver-preference-based eco-driving optimisation using a validated, forward-facing electric vehicle simulation model. The study further investigates the impact of drivetrain regenerative braking assumptions. An all-wheel-drive (AWD)-based formulation, which assumes distributed regenerative braking, is applied to a front-wheel-drive (FWD) vehicle and compared with an FWD-specific formulation that explicitly enforces front-axle regenerative limits within the optimisation. The results show that both optimisation formulations reduce energy loss relative to the IDM baseline; however, the FWD-specific formulation achieves consistently greater reductions. The generalised AWD-based formulation produces braking demands that exceed the feasible regenerative capability of the FWD drivetrain, leading to increased reliance on mechanical braking and higher electrical losses. For equivalent car-following performance, the drivetrain-inconsistent regenerative braking assumption results in 3%–11% higher total energy losses in urban driving and 7%–12% in rural conditions, with smaller differences (2%–4.1%) in mixed cycles dominated by high-speed operation.