Modélisation et validation des modèles de véhicules hybrides : étude de cas de la Toyota Prius

The objective of this thesis is to evaluate the potential of vehicle fuel economy improvement offered by thermal/electric hybrid power trains. Several configurations are first investigated. Modeling and energy consumption comparison are realized of 4 hybrid power trains showing similar performance to a reference vehicle. The modeling process carries the design of the electromechanical transmission as well as the energy management supervisory controller (EMS). The series/parallel configuration has been identified as the most promising hybrid power train technology, through its power-split eCVT transmission. Next, a thorough analysis of series/parallel powertrain is carried out: case study of the Prius. A model and an experimental validation of the 2nd generation Toyota Hybrid System power train (THS-II) of the Prius are elaborated. A rule-based (RB) EMS of THS-II is also developed. Results fully explore the potential of the power-split series/parallel powertrain in enhancing the powertrain efficiency and reducing the fuel consumption. The validation process is carried out through road test measurements realized on a MY06 Prius. The effect of the auxiliaries and the driver's attitude on the fuel consumption is also investigated. A predictive EMS for the THS-II powertrain using Dynamic Programming (DP) is also elaborated, aiming to optimize the fuel consumption over an overall specified route instead of a specific instant optimization as the RB EMS, and to accelerate the computation time of the DP algorithm in order to move a step forward toward implementing this controller into HEV real-time applications. The predictive DP EMS is assessed, compared to the developed RB EMS