Destination-Reserve-Aware Eco-MPC for Autonomous Battery Electric Vehicles
Short-horizon ecological model predictive control (Eco-MPC) coordinates energy use and car-following performance for battery electric vehicles (BEVs), whereas a destination-reserve target imposes a trip-level energy requirement. This paper proposes a destination-reserve-aware Eco-MPC that feeds remaining-trip energy adequacy into short-horizon car-following control through an outer supervisor. The supervisor blends prior and observed energy intensity to estimate remaining demand and activates a distance-indexed net-energy budget under reserve risk. A bounded energy account carries short-term budget deviations across updates. SUMO experiments include deterministic mechanism and remaining-demandestimation sensitivity analyses and paired held-out stochastic comparisons with fixed-weight and reserve-triggered adaptive-weight Eco-MPCs. At the lowest tested initial-energy level, deterministic trip energy decreased by 5.0% with a 1.2-s trip-time increase, improving the terminal reserve margin from -38.7 to -0.32 Wh. At the same initial-energy level, across three stochastic traffic conditions, mean terminal reserve margin increased by 27.2–45.5 Wh relative to fixed-weight control. Energy intensity decreased by 5.0–8.3 Wh/km, while trip time increased by 3.7–5.6 s. Relative to the specified adaptive-weight comparator, mean energy intensity decreased by 2.0–4.2 Wh/km and trip time shortened by 1.9–3.1 s. Controller differences generally diminished as initial energy increased, indicating that the proposed reserve regulation responds to the level of available trip energy.
Authors
- Jun Ma (ORCID: https://orcid.org/0000-0002-9405-8232)
- Xiao Zhang (ORCID: https://orcid.org/0000-0001-6864-2921)
- Qiumeng Li
Publication Details
- Journal
- Unmanned Systems
- Published
- 2026-11-12
- DOI
- https://doi.org/10.1142/s2301385028500847
- Primary Topic
- Traffic control and management
- Type
- article
- Field-Weighted Citation Impact
- 0.00