Automated physical realization of input, output, and compound split hybrid configurations

Abstract

Many prior studies on power-split hybrid electric vehicles focused on finding optimal configurations. Their output is usually either an optimal powertrain configuration, i.e. a diagram illustrating the connections between the four components and the planetary gear set(s) (PG), or an optimal virtual compound lever design, i.e. a bar with four nodes to which the four powertrain components are connected. Nevertheless, the actual depiction of the selected designs, which is indispensable for their physical implementation, has not yet been extensively studied. This thesis introduces the physical realization process of input and output split hybrid configurations using a single PG and compound split configurations using two PGs. First, if the selected design is a virtual compound lever, an automated methodology was developed to convert it into feasible powertrain configurations. Then, feasible kinematic diagrams, which specify the connections, locations, and arrangements of the components, are systematically generated for the obtained or selected powertrain configurations. Throughout this study, the entire design space is fully explored and the best optimal designs are selected. Multiple feasible powertrain configurations were generated for given optimal compound lever design. Some of the obtained feasible optimal powertrain configurations are not registered as patents despite their world-class performance. Then, through generating the kinematic diagrams, the best feasible components arrangements, which are good in terms of packaging and manufacturability, are selected. The generation of kinematic diagrams played a key role in choosing the best compound split configurations. In fact, many optimal compound split designs did not have good kinematic diagrams, and thus, were discarded, while many novel optimal designs were found.