Skip to Main content Skip to Navigation
Theses

Véhicule hybride et commande optimale

Abstract : In the actual automobile market, tightly connected to the overall will to decrease CO2 emissions, hybrid vehicles seem to be a necessary short and medium-term milestone. Typically, a hybrid vehicle has two energy sources: a regular gas engine is usually the main energy source, and an electric motor as a secondary energy source. The ability of a hybrid vehicle to consume less fuel and to emit less CO2 comes from its electric motor. It can be used either alongside with the engine, or alone, with no gas being burnt. Because of the two energy sources, the global powertrain needs to be managed by a strategy that chooses the power split between the engine and the motor, based on the state of charge of the battery. This power split can be optimized based on criteria like fuel consumption, CO2 emissions, etc. The objective of this thesis is to develop optimization models on optimal torque repartition between the motor and the engine of a hybrid vehicle, in order to minimize the CO2 emissions. The first step is to develop models corresponding to a reference hybrid powertrain, which are adapted to the optimization algorithms. The optimization algorithms are different depending if offline or real-time optimization is considered. Among the studied offline algorithms, dynamic programming was used to determine the optimal size of the main elements of the powertrain, and to determine the optimal gain compared to a traditional vehicle. Additionally, an original shooting algorithm called SCOP has been developed, and is able to treat optimal control problems with state constraints. This algorithm is about 50 times faster than dynamic programming. A real-time control strategy, based on the Equivalent Consumption Minimization Strategy (ECMS) and using Pontryagin's Principle, was developed and used on a hybrid prototype vehicle: a Smart equipped with a starter alternator. The results demonstrate the positive impact of the strategy to reduce the fuel consumption and the CO2 emissions.
Document type :
Theses
Complete list of metadatas

Cited literature [62 references]  Display  Hide  Download

https://pastel.archives-ouvertes.fr/pastel-00005237
Contributor : Ecole Mines Paristech <>
Submitted on : Tuesday, June 23, 2009 - 8:00:00 AM
Last modification on : Wednesday, October 14, 2020 - 3:52:33 AM
Long-term archiving on: : Friday, September 10, 2010 - 1:42:27 PM

Identifiers

  • HAL Id : pastel-00005237, version 1

Citation

Grégory Rousseau. Véhicule hybride et commande optimale. domain_other. École Nationale Supérieure des Mines de Paris, 2008. Français. ⟨NNT : 2008ENMP1604⟩. ⟨pastel-00005237⟩

Share

Metrics

Record views

1943

Files downloads

14655