Skip to Main content Skip to Navigation
Theses

Stochastic optimal control for the energy management of hybrid electric vehicles under traffic constraints

Abstract : The focus of this PhD thesis is to design an optimal Energy Management System (EMS) for a Hybrid Electric Vehicle (HEV) following traffic constraints.In the current state of the art, EMS are typically divided between real-time designs relying on local optimization methods, and global optimization that is only suitable for off-line use due to computational constraints.The starting point of the thesis is that in terms of energy consumption, the stochastic aspect of the traffic conditions can be accurately modelled thanks to (speed,acceleration) probability distributions.In order to reduce the data size of the model, we use clustering techniques based on the Wasserstein distance, the corresponding barycenters being computed by either a Sinkhorn or Stochastic Alternate Gradient method.Thanks to this stochastic traffic model, an off-line optimization can be performed to determine the optimal control (electric motor torque) that minimizes the fuel consumption of the HEV over a certain road segment.Then, a bi-level algorithm takes advantage of this information to optimize the consumption over a whole travel, the upper level optimization being deterministic and therefore fast enough for real-time implementation.We illustrate the relevance of the traffic model and the bi-level optimization, using both traffic data generated by a simulator, as well as some actual traffic data recorded near Lyon (France).Finally, we investigate the extension of the bi-level algorithm to the eco-routing problem, using an augmented graph to track the state of charge information over the road network.
Document type :
Theses
Complete list of metadatas

Cited literature [94 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02443292
Contributor : Abes Star :  Contact
Submitted on : Friday, January 17, 2020 - 9:14:42 AM
Last modification on : Wednesday, September 2, 2020 - 3:38:10 AM

File

83182_LE_RHUN_2019_archivage.p...
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02443292, version 1

Collections

Citation

Arthur Le Rhun. Stochastic optimal control for the energy management of hybrid electric vehicles under traffic constraints. Optimization and Control [math.OC]. Université Paris Saclay (COmUE), 2019. English. ⟨NNT : 2019SACLX084⟩. ⟨tel-02443292⟩

Share

Metrics

Record views

331

Files downloads

205