Convegni ATI - Accesso riservato soci CTI

Autore: T. Donateo, L. Serrao, G. Rizzoni

Collana: CA - 62 - Fisciano 2007

Note:
In the present investigation multi-objective genetic algorithms and MCDM techniques were applied to optimization of the architecture and the control of a Hybrid Electric Vehicle (HEV) in order to reduce fuel consumption and improve performance. The investigation refers to a heavy-duty military vehicle and is included in a project funded by the U.S.Army TACOM Command.
The vehicle was simulated with a Matlab-Simulink model, while the optimization was performed with the multi-objective optimization code GA-CREA. The MCDM techniques implemented in the ModeFrontier optimization environment were used for the choice of the final optimal configurations.
Nine attributes of performance and fuel consumption evaluated with respect to seven driving cycles were initiallyconsidered as optimization goals. However, the results of a full space exploration of the problem showed that the optimization goals could be clustered in two generalized index. In this way, a two-dimension Pareto front was obtained. Among the optimal solutions of the Pareto front,few final configurations were chosen with the aid of Multi- Criteria Decision Making techniques.