Article Dans Une Revue Control Engineering Practice Année : 2026

Experimental validation of predictive optimal control based eco-driving

Résumé

The main goal of Eco-Driving (ED) is to maximize energy efficiency. The paper's main novelty is a real-world evaluation of an optimal control based ED advisory system for electric vehicles, derived using Pontryagin's Minimum Principle (PMP). A Visual Advisory System (VAS), such as a digital tablet, advises the driver to follow a target eco-speed in real-time. Tests were performed with two Renault Zoe vehicles in real world traffic conditions, one equipped with the ED system and one standard, over three routes around Rueil-Malmaison, France. The ED vehicle achieved average energy savings of 4.6 % on the regional route, 13.5 % on the urban route, and approximately 4.0 % on the highway.

An analysis, using Dynamic programming (DP) as benchmark, is performed to quantify the impacts of modeling assumptions and driver behavior. Segment-level analysis shows that model simplifications in the PMP formulation account for up to a 12.6 % increase in energy consumption compared to DP, while the driver tracking error contributes an additional 0.5 % in unconstrained scenarios. In contrast, in constrained driving situations involving a preceding vehicle, driver tracking error alone leads to an energy increase of 11.3 % with respect to the advised speed profile, highlighting the influence of traffic on speed tracking capability.

Fichier principal
Vignette du fichier
Experimental validation.pdf (4.99 Mo) Télécharger le fichier
Origine Publication financée par une institution
Licence

Dates et versions

hal-05590826 , version 1 (14-04-2026)

Licence

Identifiants

Citer

Vinith Kumar Lakshmanan, Antonio Sciarretta. Experimental validation of predictive optimal control based eco-driving. Control Engineering Practice, 2026, 173, pp.106932. ⟨10.1016/j.conengprac.2026.106932⟩. ⟨hal-05590826⟩

Collections

10 Consultations
1 Téléchargements

Altmetric

Partager

  • More