Systematic elaboration of online energy management laws for hybrid vehicles
Résumé
In this paper, we describe a systematic methodology to elaborate hybrid vehicle energy management laws. "Systematic" means that this conception is done with minimal tuning, whatever the hybrid vehicle architecture and sizing are. A radial basis function type neural network is used to learn from the offline energy management laws given by an optimization tool named KOALA. Then the trained neural network can be used to control the hybrid vehicle online. The results are satisfactory as regards hybrid vehicle fuel consumption. This study has been implemented on a high dynamic test bench in hardware in the loop, and is still in progress.