Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


3133.pdf (2.25 MB)

ETD Abstract Container

Abstract Header

Evaluation of Embedded Prognostic Design in High Dimensional Data Environment

Jiang, Chuan
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1382372538

Abstract Details

2013, MS, University of Cincinnati, Engineering and Applied Science: Mechanical Engineering.
In the past decades, numerous efforts have been focused on the development of prognostic algorithms in the area of Prognostic and Health Management (PHM). Today’s engineering systems have achieved increased safety, reliability and sustainability. However, it has been observed that many prognostic algorithms are not compatible with embedded systems because of limited computational resources, data transferring speed and storage space. These problems obstructed a lot for the implementation of embedded prognostic applications. To deal with this issue, a systematic architecture of embedded prognostic application design and the main procedures for embedded prognostic applications are illustrated. Also, a simple principal component analysis (SPCA) for feature dimensionality reduction is introduced for embedded prognostic applications in the thesis. Compared with principal component analysis (PCA) and kernel principal component analysis (KPCA), SPCA is not only an effective dimension reduction algorithm, but also has better performance of computation speed and adaptive capability in online embedded prognostic applications. The proposed methodology is validated in two applications, a rolling element bearing test and an imbalanced shaft fault diagnosis with algorithms deployed on an embedded system. Results show that the calculation of SPCA is much faster than PCA and KPCA when running in embedded systems, especially for high dimensional data set. In addition, it can be adaptive to unknown faults of the machine so the fault detection accuracy is enhanced. This merit also helps to save data storage space for embedded systems. Moreover, the reasons why SPCA fits better for embedded systems are discussed theoretically and thoroughly from the perspective of mathematical steps. It provides insight and guidance to develop new algorithms or modify current algorithms which are more compatible for embedded prognostic applications.
Jay Lee, Ph.D. (Committee Chair)
Hongdao Huang, Ph.D. (Committee Member)
Mark Schulz, Ph.D. (Committee Member)
74 p.
Mechanics

Recommended Citations

Citations

  • Jiang, C. (2013). Evaluation of Embedded Prognostic Design in High Dimensional Data Environment [Master's thesis, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1382372538

    APA Style (7th edition)

  • Jiang, Chuan. Evaluation of Embedded Prognostic Design in High Dimensional Data Environment. 2013. University of Cincinnati, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1382372538.

    MLA Style (8th edition)

  • Jiang, Chuan. "Evaluation of Embedded Prognostic Design in High Dimensional Data Environment." Master's thesis, University of Cincinnati, 2013. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1382372538

    Chicago Manual of Style (17th edition)

ucin1382372538
324
© 2013, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.