Validation of robustness and fuel efficiency of a universal model-based energy management strategy for fuel cell hybrid trains: From analytical derivation via simulation to measurement on test bench

•Implementation of offline PMP considering accurate battery modeling as references.•Derivation of an analytical formula to estimate costate from energy conservation.•Validation of the robustness of APMP against various conditions of uncertainty.•Experimental validation of APMP regarding fuel efficiency and robustness.•A universally applicable model-based energy management strategy. Fuel cell hybrid trains are being commercialized to replace trains powered by combustion engine to reduce carbon dioxide emission without high investment cost in overhead catenaries. In this context, this paper presents a universal model-based strategy for the operation of fuel cell hybrid trains based on adaptive Pontryagin’s minimum principle (APMP). Different from all other work, the implementation of Pontryagin’s minimum principle (PMP) considers the relaxation process due to the resistance-capacitor branches in the batteries to provide a precise reference for the evaluation of the robustness and fuel economy of the APMP-based strategy. Furthermore, a formula to physically estimate the costate is inspired by the offline PMP results and derived by using the energy conservation principle. Moreover, the robustness of the strategy against fuel cell aging, battery aging, inaccurate fuel cell modeling, and deviations introduced through fitting experimental data is investigated through simulation. Compared to the offline results, a maximum 1.5% higher hydrogen consumption is observed by simulation under different aging and uncertain operating conditions. Finally, the effectiveness and the robustness of the strategy are validated through measurement on the test bench at the Center for Mobile Propulsion of the RWTH Aachen University. A maximum of 2.7% more hydrogen consumption is measured compared to the offline PMP results under various conditions of uncertainty.