Hybridization of a vehicle’s drivetrain can in principle help to improve its energy efficiency by allowing for recuperation of kinetic energy and modulating the engine’s load. How well this can be realized depends on appropriate sizing and control of the additional components. The system is typically designed sequentially, with the hardware setup preceding the development and tuning of advanced controller architectures. Taking an alternative approach, component sizing and controller tuning can be addressed simultaneously through simulation-based optimization.
The results of such optimizations, especially with standard algorithms with continuous design variable ranges, can however be difficult to realize, considering for example limitations in available components. Furthermore, drive-cycle based optimizations are prone to cycle-beating. This paper examines the results of such simulation-based optimization for a series hydraulic hybrid vehicle in terms of sensitivity to variations in design parameters, system parameters and drive cycle variations. Additional relevant aspects concerning the definition of the optimization problem are pointed out.