Robust Combined Design and Control Optimization of Hybrid-Electric Vehicles Using MDSDO

Abstract

Identifying a true, system-level optimal solution for hybrid-electric vehicle (HEV) powertrains requires methods such as combined design and control optimization (co-design) to account for the coupling between powertrain design and optimal energy management strategies. Despite the widespread application of co-design techniques to HEV powertrains in recent years, limited research has been done to account for the inherent uncertainties present in these systems. This is problematic as such uncertainties may not only impact HEV powertrain component sizing and control strategies, but also overall vehicle performance and cost-both of which are critically-important in the fiercely-competitive automotive industry. One way to address this issue is to apply a stochastic dynamic system optimization technique known as robust multidisciplinary dynamic system design optimization (R-MDSDO) to the HEV powertrain co-design problem. This method ensures that the HEV powertrain co-design solution, including its system performance and cost, explicitly accounts for random variations within the system and its problem formulation using concepts from robust design optimization. Therefore, in this study, we formulate and solve a robust co-design problem for a power-split HEV powertrain using R-MDSDO to identify the optimal component designs, state trajectories, and control trajectories such that the vehicle powertrain cost is minimized. The robust solution is then compared to the solution from the associated deterministic problem, with the results indicating that accounting for system uncertainty has a significant impact on HEV powertrain co-design.