Using Quasi-Newton method for energy management in electrical multi source systems

Finding a compromise for the use of several energy sources in order to meet the power demand of an electrical powertrain system is the objective of energy global optimization. The challenge is to design more efficient algorithms that give better result on consumption costs and require less computing time. The purpose remains to minimize the criterion “hydrogen consumption” whilst satisfying the power demand of the powertrain, by finding an optimal energy splitting strategy between the multi-electrical sources. New heuristics presented in the paper are based on a global optimization approach, have a minimum computing time and more importantly a lower cost consumption than the solutions from the literature, even dynamic programming. Finally, a lower bound of the optimal consumption has been computed and helps to better evaluate the quality of the different solutions proposed.