Rotating Machinery Prognostics and Application of Machine Learning Algorithms Use of Deep Learning with Similarity Index Measure for Health Status Prediction

##plugins.themes.bootstrap3.article.main##

##plugins.themes.bootstrap3.article.sidebar##

Published Sep 24, 2018
Asheber Wagshum Techane Yu-Fu Wang Bereket Haile Weldegiorgis

Abstract

The internet of things (IOT) enabled presence of abundant sensors on smart machineries and the recent advance in deep learning is accelerating the development of predictive maintenance in production systems with less time and fair amount of effort. In this work a Deep learning Neural Networks (DNN) based bearing health monitoring system with index of similarity check is developed and tested for its effectiveness. The assessment procedure followed in here trains a DNN model on a time series data segmented to a vector size equal to number of data points per cycle as training and test data sets. Moreover the model measures the similarity of the test signal to an anchor signal selected from each fault class. The classification performance comparison done proves that DNN with fair depth and large data perform better and can be extended to other problems in intelligent maintenance systems development efforts. The proposed system is intuitive and has minimal complexity with uncompromised fault detection accuracy.

How to Cite

Techane, A. W., Wang, Y.-F., & Weldegiorgis, B. H. (2018). Rotating Machinery Prognostics and Application of Machine Learning Algorithms: Use of Deep Learning with Similarity Index Measure for Health Status Prediction. Annual Conference of the PHM Society, 10(1). https://doi.org/10.36001/phmconf.2018.v10i1.253
Abstract 882 | PDF Downloads 520

##plugins.themes.bootstrap3.article.details##

Keywords

DNN, times series data, similarity index

Section
Technical Papers