Please use this identifier to cite or link to this item:
Title: Fault diagnose based on pattern recognition
Authors: Liu, Zhuoshi.
Keywords: DRNTU::Engineering
Issue Date: 2013
Abstract: This is a joint project with SIMTech. It is mainly focusing on developing a fault diagnose system. The system can be used to solve stochastic and dynamical problems, like bank abnormal transaction detection, operation abnormal event detection and equipment failure event detection. This system firstly takes training data, which is consisted of multi-dimensional input x and output y, and the parameters inside the system learn from the training data by minimizing the differences between predicted value and real output y. Then, the system will take testing data and predict the output using the trained parameters inside the system. Throughout the calculation, “Kernel recursive least square” (KRLS) method acts as the most important part. The KRLS algorithm presents a nonlinear version of the recursive least squares (RLS) algorithm. The algorithm performs linear regression in a high-dimensional feature space induced by a Mercer kernel and can therefore be used to recursively construct minimum mean-squared-error solutions to nonlinear least-squares problems that are frequently encountered in signal processing applications. In a fault diagnose system, the accuracy of prediction is one of the most important part. Thus, in this project, the author will focus on improving the accuracy of the system based on the partially implemented system by the previous student. Besides, the author also enhanced the user interface, to make it more user-friendly.
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
  Restricted Access
2.28 MBAdobe PDFView/Open

Page view(s)

Updated on Dec 3, 2020

Download(s) 10

Updated on Dec 3, 2020

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.