 |
 |
| |
Selected Development Project |
| |
 |
| Project Title |
Advanced Statistical Models for Accelerated Life Testing Data of One-shot Devices
對一次性使用的裝置的加速壽命試驗數據的統計模型 |
|
| |
| Principal Investigator |
Dr LING Man Ho |
|
| |
| Area of Research Project |
Mathematics and Information Technology |
|
|
|
|
|
| |
| Project Period |
|
|
| Objectives |
- To develop statistical models and statistical methods to help engineers and scientists analyze one-shot device testing data collected from accelerated life testing
- To discriminate among common lifetime distributions, namely log-normal, Weibull, gamma distributions, and to evaluate the impact on experimental design when an inappropriate statistical model is selected to represent the lifetime of devices
- To develop a framework to design cost-effective accelerated life testing for collecting lifetime information of one-shot devices
- To test and examine the developed algorithms using real examples, simulation studies, and to provide practical advice to engineers, experiment designers, and scientists
|
|
|
| Methods Used |
We plan to test the proposed methods for modeling one-shot device testing data with engineering case study as well as simulation studies. We also plan to conduct sensitivity analysis for evaluating the misspecification effects of inference on the mean lifetime as well as optimal test planning for different lifetime distribution models.
|
Impact |
The proposed research will help us to obtain accurate reliability estimations based on accelerated life testing data of one-shot devices. These precise reliability estimations will also allow engineers and scientists to devise a cost-effective maintenance schedule as well as to design cost-effective life testing for one-shot devices. Another major impact of this research is that it would effectively identify key covariates for reliability improvements in analyzing binary response data in various important scientific applications. In addition, the effect of model-misspecification on inference will be investigated in this project. This will help engineers and scientists to understand the impact of model misspecification on reliability estimation. Our research will try to develop efficient statistical algorithms for computation, effective statistical methods for model validation, and advanced statistical methodologies and techniques for data collection in various applications. |
Selected Output |
- Ling, M.H., Ng, H.K.T., Chan, P.S., Balakrishnan, N. Autopsy data analysis for a series system with active redundancy under a load-sharing model, IEEE Transactions on Reliability, 65, 957-968, 2016.
- Ling, M.H., Ng, H.K.T., Tsui, K.L. Inference on remaining useful life under gamma degradation models with random effects, In Statistical Modeling for Degradation Data, Springer, accepted.
- Ling, M.H., Balakrishnan, N. Model mis-specification analyses of Weibull and gamma models for one-shot device testing data, IEEE Transactions on Reliability, in press.
|
| Biography of Principal Investigator |
Dr LING Man Ho is an Assistant Professor in the Department of Mathematics and Information Technology, Faculty of Liberal Arts and Social Sciences. He received his B.Sc. and M.Phil. degrees from Hong Kong Baptist University, Hong Kong, in 2005 and 2008, respectively. He obtained his Ph.D. degree from McMaster University, Hamilton, Ontario, Canada, in 2012. Dr Ling’s research interests include reliability and survival analysis, binary data, statistical inference under censoring and statistical computing. |
Funding Source |
General Research Fund |
|
|
 |
|
 |
