Optimization of hybrid electric cars by neuro-fuzzy networks

In this paper, the problem of the optimization of energetic flows in hybrid electric vehicles is faced. We consider a hybrid electric vehicle equipped with batteries, a thermal engine (or fuel cells), ultracapacitors and an electric engine. The energetic flows are optimized by using a control strategy based on the prediction of short-term and medium-term vehicle states (energy consumption, vehicle load, current route, traffic flow, etc.). The prediction will be performed by a neuro-fuzzy control unit, where the predictive model exploits the robustness of fuzzy logic in managing the said uncertainties and the neural approach as a data driven tool for non-linear control modeling.