Application of frequency domain experiments in the factor screening of production systems

The purpose of this study was to demonstrate that the Application of Frequency Domain Experiments (AFDE) method is effective in screening for discrete variable factors within production systems, plus requires fewer computer simulations than conventional factorial design. To help clarify this, an ass...

Full description

Saved in:
Bibliographic Details
Main Author: Minsan,W.
Format: Article
Published: Chiang Mai University 2015
Subjects:
Online Access:http://www.scopus.com/inward/record.url?partnerID=HzOxMe3b&scp=84860736764&origin=inward
http://cmuir.cmu.ac.th/handle/6653943832/38628
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Chiang Mai University
Description
Summary:The purpose of this study was to demonstrate that the Application of Frequency Domain Experiments (AFDE) method is effective in screening for discrete variable factors within production systems, plus requires fewer computer simulations than conventional factorial design. To help clarify this, an assembly line production system was designed in order to have factor screening experiments carried out using the AFDE method. The study assembly line produces one type of product and there were ten workstations assigned to be the study's discrete variable factors. Each factor had one to five machines for the workers to control. In the experiment, there were three main study aims in terms of the assembly line production process, these being to carry out: 1) A Major Factor Study, 2) Impact Ranking Major Factors Study, and 3) Sensitivity Impact of the Factors Study. The study results show that AFDE had the ability to detect bottleneck workstations in the production system for all the types of study undertaken. When assigning the system to have one, two or three bottleneck stations in a respective order, AFDE was able to detect these bottleneck stations and order the importance of the stations with 100% accuracy. In addition, AFDE requires fewer computer runs than conventional factorial design.