Search Results

In the past decade, various Plug-in Hybrid Electric Vehicles have been demonstrated which offer the potential of a significant reduction in fuel consumption and emission. However, this capability strongly depends on the sizing of the components, driver's usage profile and the quality of the energy management. These challenges require new optimization procedures for a systematical exploration of the design space with the objective of an optimal powertrain configuration. A novel optimization strategy based on a multi-objective problem formulation is proposed. The optimization procedure consists of a multi-objective genetic algorithm for determining the best design parameters with respect to fuel consumption and driving performance. The approach is combined with an analytical optimal control problem to find the optimal continuous and discrete control trajectories for the energy management.

The quality of the powertrain design has a significant impact on the fuel consumption and emissions of hybrid vehicles. Lack of experience with these relatively new technologies, the enormous variety of hybrid powertrain configurations, and the multitude of components make this area an ideal application for computer-based modeling and optimizations. Global optimization techniques have the advantage to explore systematically the design space to find the optimal configuration space. In this paper, a systematic procedure for an optimal design of hybrid powertrain configurations using an evolutionary algorithm is proposed. It will be shown that the design steps for parallel and power-split configurations are quite similar. This results in a computing approach with high synergy effects and the ability to exchange components seamless to compare different ‘virtual’ configurations.