Mechanical Engineering MS Thesis Defense by Mr. Yabo Han
When: Wednesday,
September 30, 2015
2:00 PM
-
4:00 PM
Where: Textiles Building 101E
Description: Mechanical Engineering MS Thesis Defense by Mr. Yabo Han
September 30, 2015
2:00 p.m. - 4:00 p.m.
Textile Building, Room 101E
TOPIC:
Construction of a Multivariate Control Chart Based on Kernel Density Estimation and Metropolis-Hastings Sampling
ABSTRACT:
Statistical process control (SPC) is a set of procedures that uses statistical techniques to measure, analyze, and reduce process variation. Control chart is the commonly used and most successful SPC tool in real-world applications.
Traditional Multivariate Quality Control Charts (MQCC), such as Hotelling T2 control chart, Multivariate Cumulative Sum (MCUSUM) and Multivariate Exponentially-Weighted Moving Average (MEWMA) control chart, etc. are based on multivariate normal assumption. This given assumption is not usually satisfied, especially in modern industrial manufacturing process.
In this study, a new method to construct control chart for unknown multivariate distributional process is proposed, which is based on Kernel Density Estimation (non-parameter density estimation method) and Metropolis-Hastings sampling. This new method is able to avoid given assumption and accommodate non-normal or normal multivariate process. The control limit for control chart is a constant value which is derived by Metropolis-Hastings sampling method in a numerical way. Two simulation cases used to evaluate and validate the performance of the proposed method will be provided in this study.
ADVISOR:
Dr. Wenzhen Huang (whuang@umassd.edu, 508-910-6568)
COMMITTEE MEMBERS:
Dr. Farhad Azadivar and Dr. Vijaya Chalivendra
Open to the public. All MNE students are encouraged to attend.
For more information please contact:
Dr. Wenzhen Huang (whuang@umassd.edu, 508-910-6568)
Thank you, Sue Cunha, Administrative Assistant
September 30, 2015
2:00 p.m. - 4:00 p.m.
Textile Building, Room 101E
TOPIC:
Construction of a Multivariate Control Chart Based on Kernel Density Estimation and Metropolis-Hastings Sampling
ABSTRACT:
Statistical process control (SPC) is a set of procedures that uses statistical techniques to measure, analyze, and reduce process variation. Control chart is the commonly used and most successful SPC tool in real-world applications.
Traditional Multivariate Quality Control Charts (MQCC), such as Hotelling T2 control chart, Multivariate Cumulative Sum (MCUSUM) and Multivariate Exponentially-Weighted Moving Average (MEWMA) control chart, etc. are based on multivariate normal assumption. This given assumption is not usually satisfied, especially in modern industrial manufacturing process.
In this study, a new method to construct control chart for unknown multivariate distributional process is proposed, which is based on Kernel Density Estimation (non-parameter density estimation method) and Metropolis-Hastings sampling. This new method is able to avoid given assumption and accommodate non-normal or normal multivariate process. The control limit for control chart is a constant value which is derived by Metropolis-Hastings sampling method in a numerical way. Two simulation cases used to evaluate and validate the performance of the proposed method will be provided in this study.
ADVISOR:
Dr. Wenzhen Huang (whuang@umassd.edu, 508-910-6568)
COMMITTEE MEMBERS:
Dr. Farhad Azadivar and Dr. Vijaya Chalivendra
Open to the public. All MNE students are encouraged to attend.
For more information please contact:
Dr. Wenzhen Huang (whuang@umassd.edu, 508-910-6568)
Thank you, Sue Cunha, Administrative Assistant
Topical Areas: General Public, University Community, College of Engineering, Mechanical Engineering