2010 Prognostics & System Health Management Conf., Macau, China, Jan. 12-14, 2010.

Study of Ensemble Learning-Based Fusion Prognostics

Sun Jianzhong13, Zuo Hongfu1,Yang Haibin1  Michael Pecht2 3

  1. Nanjing University of Aeronautics and Astronautics, 210016, Nanjing, P.R. China.
  2. Prognostics and Health Management Center, City University of Hong Kong
  3. Center for Advanced Life Cycle Engineering (CALCE) , University of Maryland,
    E_mail: sunjianzhong@nuaa.edu.cn. Phone: 86-25-84895772  Fax: 86-25-84890647

 

Abstract:

In this paper we explore the effectiveness of ensemble learning in the failure prognosis field by MLP neural network. An effective ensemble should consist of a set of learners that are both accurate and diverse. In the training stage, we use the Adaboost.R2 technique to train several weak learners (multi layer perceptron network-MLP) to increase the diversity of the individual models. In the prediction stage, we focus on the design of the fusion of the weak learner ensemble. In contrast with the traditional static weight allocation method based on the overall data set, we propose a dynamic weight allocation method, based on the performance of an individual weak learner on the subset. The idea behind this method was to use a test sample’s neighbours to estimate the accuracy and bias of the individual weak learner when the learner is used to make the prediction. An improved hyper rectangle neighbourhood defining method is proposed in this paper. Some experiments using MLP neural network as a weak learner on a NASA turbofan engine degradation simulation dataset were carried out. The preliminary empirical comparisons showed higher performance of the novel ensemble learning methodology for the RUL estimation of engineering systems.

Index Terms: Prognostics, Health Management, Ensemble Learning, Machine Learning, Neural Network Ensemble

Complete article is available to CALCE Consortium Members.

© IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

 



[Home Page] [Articles Page]
Copyright © 2010 by CALCE and the University of Maryland, All Rights Reserved