A Methodology to Support Product Platform Optimization Using Multi-objective Evolutionary Algorithms


A critical step when designing a successful product family is to determine a cost-saving platform configuration along with an optimally distinct set of product variants that target different market segments. A multi-objective optimization-based platform design methodology (MOPDM) was presented to optimize the individual product performances with a feasible platform commonality level. The process and optimization model for scale-based product platform was constructed firstly, and then the MOPDM was carried out in two stages using the non-dominated sorting genetic algorithm II (NSGA-II). A mechanism based on fuzzy set theory was developed to extract one of the Pareto-optimal solutions as the best compromise one. During the first stage of MOPDM, each product in the family was optimized independently with NSGA-II. Those design variables that show small deviations were held constant to form the product platform. The scaling variables of each instance product were optimized in the second stage. The efficiency and effectiveness of proposed method is illustrated by optimizing a family of six capacitor-run single-phase induction motors, and the results are compared against previous work.