Plug-in hybrid electric vehicles (PHEVs) have been demonstrated as auspicious solutions for ensuring improvements in fuel-saving and emission reductions and a good transition towards full electric vehicles. Design optimization tools for powertrain systems and sub-systems play crucial roles, which can enable high-efficiency, high power-density at a concurrently lower total cost of ownership, weight, volume and less development time.
The analysis of the relevant literature shows that hardware and controller design for power electronics and automotive engineering systems have been treated independently. However, a change in hardware parameters may lead to changes in the overall system performance. Therefore, the integration of the hardware and controller design into a holistic framework, called codesign optimization framework (COF), can possibly outperform the overall outcome of the traditional sequential approach. The main focus of this PhD thesis is to develop two advanced COFs, which will be adopted differently for two use-cases: power electronics converters and PHEV powertrains.
In the first use-case, this PhD thesis will propose a novel simultaneous COF to optimize the power conversion stage and dual-loop controller of a non-isolated interleaved boost converter in a SiC-based multiport converter. The COF adopts a non-dominated sorted genetic algorithm (NSGA-II) and average ranking technique to solve a multi-objective optimization problem including four objective functions. Based on the optimal solution, a 60kW prototype of SiC-based MPC has been built and tested in the laboratory to verify the proposed simultaneous COF.
In the second case-study, this PhD research will introduce a new nested COF based on forward-facing model for the PHEV powertrain of a heavy-duty distribution truck. COF consists of an inner loop (i.e. an optimal control strategy using an equivalent consumption minimization strategy), which is nested inside an outer loop (i.e. a component sizing optimization loop employing genetic algorithms).