Electrothermal Dynamics-Conscious Many-Objective Modular Design for Power-Split Plug-in Hybrid Electric Vehicles

This article proposes an improved modular design methodology of a power-split plug-in hybrid electric vehicle (PHEV) that introduces an advanced electrothermal coupled model and a temperature-related subobjective to simultaneously reveal battery thermal and electrical dynamics in the modular design. Considering to provide customers with more optimal configuration solutions, a Pareto-augmented collaborative optimization (PACO) scheme is designed that integrates three benchmarking many-objective evolutionary algorithms (MOEAs) to expand the distribution of an approximated Pareto frontier composed of the best solution set. Two realistic worldwide harmonized light vehicles test cycles are separately reproduced by two trained drivers on a chassis dynamometer to test the robustness of the optimized vehicle system. The simulation results demonstrate that the MOEA based on decomposition (MOEA/D) in the PACO is the main contributor for PHEV modular design because it lessens the generational distance by at least 2.7% and enlarges the hypervolume by at least 17.6%, compared to the elitist nondominated sorting genetic algorithm and improved strength Pareto evolutionary algorithm. In the modular adaptation for different user types, the PHEV system optimized by the PACO can regulate cell temperatures (
$\mathbf{27}{{\bf .5}} - \mathbf{38}{{\bf .}}{\mathbf{3}^ \circ }\mathrm{C}$
) of all user types within a safe and efficient working zone (
$\mathbf{0} - \mathbf{5}{\mathbf{5}^ \circ }\mathrm{C}$
).