Development of a Desirability Method for Optimization of a Sizing System based on Anthropometric Data

Abstract

A good sizing system needs to anthropometrically accommodate a design target population with acceptable fitness and efficiency; however, it is a great challenge to establish a sizing system while achieving these requirements at the same time using previous methods. This study developed a new multivariate optimization method (desirability method) that can simultaneously optimize a sizing system in terms of accommodation percentage (anthropometric coverage), loss score (size fitness penalty), and overlap area (sizing system inefficiency). Performances of the proposed desirability method were compared with a representative existing method in a case study for different numbers (5, 10, and 15) of size category. The desirability method outperformed by 4% in accommodation percentage than the existing method across all size categories. On the other hand, the desirability method showed slightly larger loss scores by 1.15 and 0.61 for the small size categories (5 and 10) than the existing method. Lastly, the desirability method resulted a smaller overlap area by 5.17 ㎠ than the existing method for the largest size category considered in this study.