Metamodel-based Product Family Design Optimization for Plug-In Hybrid Electric Vehicles

Plug-in Hybrid Electric Vehicles (PHEVs) have been recognized as a solution to mitigate the green-house emission for transportation. A factor to succeed in the marketplace is to provide products that can meet customer expectations and satisfy various functional requirements. As such, the design of PHEVs for diverse market segments requires sufficient differentiation in this product to maximize customer satisfaction with the new technology. However, there are challenges coupled with diversity in production of such a complex product for various customers. This dissertation attempts to address the challenges. This thesis proposed the use of product family design to ensure both the manufacturing efficiency and the customer satisfaction for PHEVs in various market segments. A thorough review of the developments in product family design is first performed, and directions for developing an efficient family design methodology are identified. In order to select the desired or the most preferred variants for the family design purposes, a review of the market studies and fleet data for PHEVs has been performed and summarized as well, based on which a set of five PHEVs- known as variants- are selected for family design assessments. Thirdly, a methodology is proposed for PHEV product family design to enable scale-based design of the selected PHEV variants. The proposed method is verified through a test problem from the literature, and its application to the PHEVs design provides design solutions for the PHEV product family under study. Since the vehicle performance is assessed through expensive simulations, it is shown that the selected optimization algorithm, along with the commonalization strategy and the decision criteria for commonalizing specific design variables make an efficient methodology in terms of the computational costs, and the overall performance of the obtained family solutions. The proposed methodology can also find applications in other product designs that involve expensive simulations and unknown design equations.