Emmanuel Vinot

Improvement of an EVT-Based HEV Using Dynamic Programming

Automotive engineers and researchers have proposed different topologies for series-parallel hybrid electric vehicles (SP-HEVs). The Toyota Hybrid System (THS) is the best known SP-HEV-based vehicle, but alternative solutions, such as electric variable transmission (EVT), have been also proposed. An efficient comparison between these different solutions is a key point to estimate the added value of each topology. This paper presents the application of optimal control to two series-parallel hybrid architectures for efficiency assessment purposes. The dynamic programming method is applied to the THS and to a virtual hybrid vehicle with an EVT. The way to take into account the supplementary degree of freedom provided by the decoupling of the wheels and the engine in both topologies is presented. The optimal fuel consumption is then compared on different driving cycles and brings out an overconsumption of the EVT topology. Then, a parametric study shows that inserting an appropriate gear ratio on the internal-combustion-engine (ICE) shaft can improve the EVT efficiency that becomes close to the THS efficiency.

Model simulation, validation and case study of the 2004 THS of Toyota Prius

For hybrid vehicles, it is difficult to obtain realistic component models and actual interaction between subsystems without carrying out experiments. INRETS has performed a coupled approach associating simulation using its in house VEHLIB software together with experimental validation. This paper highlights the evaluation and modelling of the Toyota Hybrid Systems (THS) used in the 2004 Prius vehicle. First, the ICE, battery, boost converter, and strategy evaluation and validation process to obtain a useful model of the vehicle are examined. Then examples of application are presented concerning energy flow in the vehicle and engine operation compared to a conventional car.

Modeling power semiconductor losses in HEV powertrains using Si and SiC devices

Silicon carbide (SiC) power semiconductor devices are known to have potential benefits over conventional silicon (Si) devices, particularly in high power applications such as hybrid electric vehicles (HEVs). Recent literature studying the use of SiC JFETs in HEV inverters indicate a substantially increased gas mileage. This paper further investigates this change in inverter efficiency due to the adoption of SiC using analytical loss models and empirical loss data obtained from experimental Cree 1200V 10A DMOSFETs and Schottky diodes. A motor inverter efficiency map is developed and used in the VEHLIB simulator to evaluate fuel consumption benefits. Distribution of conduction and switching losses in both Si and SiC inverters is explored.

HEVs Comparison and Components Sizing Using Dynamic Programming

This paper presents a backward approach for HEVs modelling with a global optimization method. This method allows to determine the maximal potential gain in terms of CO2 emission for a given HEV. It is then applied to size and compare two hybrid architectures and conventional vehicles. Finally a comparison on measurement, classical forward approach and optimized backward approach is performed on the Toyota Prius 2004 vehicle.

Optimal management of electric vehicles with a hybrid storage system

This paper presents a comparison between two offline optimisation methods for energy management applied to electrical vehicle with one electrical machine and fed by a hybrid storage system composed of batteries and ultra-capacitors. After a short presentation of the two methods, they are applied and compared to the case of an electric micro bus.

Global Optimized Design of an Electric Variable Transmission for HEVs

A way to improve existing hybrid vehicles is by globally optimizing the design of their components in relation to their actual use. This paper proposes a global optimal design method for power-split hybrid electric vehicles (PS-HEVs) equipped with an electric variable transmission (EVT). A genetic algorithm is used to optimize system parameters. The process includes discrete dynamic programming (DDP) for optimal energy management. The optimization focuses on minimizing fuel consumption and the number of battery cells. Pareto fronts are depicted and compared for different drive cycles, and they show the existence of a possible tradeoff between fuel consumption and battery size.

Global optimization of a parallel hybrid vehicle using optimal energy management

Fuel consumption of a hybrid electrical vehicle (HEV) depends on the driving cycle (i.e. the use of the vehicle), on the powertrain components and on the energy management. The paper proposes an optimal component sizing process taking into account the entire system. An optimal energy management is used in order to minimize the fuel consumption on a driving cycle. This optimization approach of the HEV uses Radial Basis Function models which allow a fast constrained optimization.

A common model validation in the case of the Toyota Prius II

Using different approaches, modeling for different HEVs has been studied. Generally, considering the different combinations of components, each architecture has its own modeling. The modeling and control design of different HEVs could be achieved in a general way, despite the fact that HEVs can be very different from each other in terms of structure. If the modeling and control design of different HEVs is obtained using a general approach, the time required for control design can be reduced. A common simulation program has been established for different HEVs. By setting different values for certain parameters, this single simulation program can be used for series-parallel, series, and parallel HEVs. The objective of this paper is to validate this common model in the case of the Toyota Prius II. The simulation validation is conducted by comparing it with VEHLIB (Hybrid Electric Vehicle Library), a simulation software program, which has already been validated by a real Toyota Prius II.

Optimal Management and Comparison of SP-HEV vehicles using the dynamic programming method

This paper presents the application of optimal control to two series-parallel hybrid architectures. The dynamic programming method is applied to Toyota Hybrid System (THS) and a hybrid vehicle with Electric Variable Transmission (EVT). First the theoretical approach of dynamic programming and its application to SP-HEVs is presented. The way to take into account the extra degree of freedom, provided by the decoupling of wheel and engine, is presented. Then, the optimal fuel consumption of the two architectures is compared on different driving cycles for the case of a same reference vehicle Prius II. The over-consumption of the EVT topology is analysed and explained.

Energy management system for hybrid electric vehicle: Real-time validation of the VEHLIB dedicated library

This paper deals with the energy share between batteries and supercapacitors within hybrid electric vehicles (HEV). A library of models, known as Hybrid Electric Vehicle Library (VEHLIB), which combines the different models to form a coherent modular base, has been constructed and implemented in real time simulator. Real-time results are here discussed in order to illustrate the effectives of models used. The integration of the on-board energy source of an electrically propelled vehicle with a supercapacitor bank (SB) as a peak power unit, can lead to substantial benefits in terms of electric vehicle, battery life and energy economy. An energy management strategy is also suggested, which is based upon the DC bus voltage regulation.