Using Multivariate Analysis to Select Accommodation Boundary Manikins from a Population Database
Open Access
- Author:
- Boyd, Devon Kylee
- Graduate Program:
- Mechanical Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- April 17, 2015
- Committee Members:
- Matthew B Parkinson, Thesis Advisor/Co-Advisor
- Keywords:
- Digital Human Models
Anthropometry
Multivariate Analysis - Abstract:
- Digital Human Models (DHMs) are a tool that can be used to aid in determining dimensions for human-centered designs. DHMs have the ability to represent the anthropometric extremes of the population and help to determine which dimensions should be used to accommodate a predefined portion of the population. In order to use DHMs within design, a subset must be determined that explains the needed variance within the target population. Current techniques such as principal component analysis (PCA), representative manikin families, and univariate analysis focus on identifying arbitrary dimensions fro the creation of manikin abstractions. A new approach for selecting mankind from an existing population is needed. The purpose of this research is to provide multivariate analyses that determines a subset of manikins representing the total target population when comparing up to eleven anthropometric dimensions of the manikins. This pool will act as boundary manikins for a given level of accommodation. The analysis when dealing with two or three possible variables is based on Pareto Optimization. Due to issues with this method being cumbersome, the analysis for four or more variables is based on the use of parallel coordinates. These methods are then made available through the use of a website application. Included within this research are case studies on the creation of groups of boundary manikins. A specific case study focuses on the importance of DHMs in design. This case study focus of the creation of a set of DHMs that can be used to aid students in the creation of competition race vehicles, such as the Formula SAE car. These methods, or tools, provide designers the opportunity to select manikins, or determine dimensions, from a predefined population dataset that meet their design needs. This selection was previously not possible for a multivariate problem. The tools allow designer to select their design manikins based on however many dimensions are important to their design. This selection will help designers to create a product that better accommodates their target population.