Multi-objective benchmark for energy management of dual-source electric vehicles: An optimal control approach

Highlights: • Two objectives of battery degradation and supercapacitors losses. • Transparent analytical form of the multi-objective optimal solutions. • State and control variables constraints fully considered. • Hundreds of times faster than dynamic programming with better results. • Applicable for both active and semi-active topologies due to general formulation. This paper proposes a novel method to develop a multi-objective optimal energy management strategy (EMS) for hybrid battery/supercapacitor (SC) electric vehicles. The method is based on an alternative approach of using Pontryagin's minimum principle (alt-PMP), which is superior to dynamic programming in terms of computational effort while obtaining better performance. The novel multi-objective EMS allocates the battery and SC powers to minimize the battery degradation and the SC subsystem losses. The proposed approach deduces transparent analytical forms of the optimal solutions, which have been rarely discussed in the literature. The results form a Pareto optimal (nondominated) front displaying the trade-offs associated with the objectives, which can serve as a benchmark to evaluate other real-time control strategies. Numerical investigations are carried out to validate the advantages of the proposed method. The benchmark role of the obtained nondominated front is illustrated by comparing it to the well-known filter-based strategy. Moreover, this study shows the conversion of the Pareto front to an "ultimate utopia point" corresponding to the ideal case of the SC subsystem efficiency. The proposed approach can be extended to dimensioning problems, to develop real-time EMS, and to more complex multi-source systems in future works.