A publication of CHEMICAL ENGINEERING TRANSACTIONS VOL. 33, 2013 The Italian Association of Chemical Engineering Online at: www.aidic.it/cet Guest Editors: Enrico Zio, Piero Baraldi Copyright © 2013, AIDIC Servizi S.r.l., ISBN 978-88-95608-24-2; ISSN 1974-9791 Diversity and Integration of Rotating Machine Health Monitoring Methods Shigeru Kanemoto* a , Norihiro Yokotsuka a , Noritaka Yusa b , Masahiko Kawabata c a The University of Aizu, Tsuruga, Ikki-machi, Aizuwakamatsu-city, 965-8580, Japan b Tohoku University, Aoba-ku, Sendai-city, 980-8579, Japan c TRIBOTEX Co., Ltd., Nagakusa-cho, Oofu-city, 474-0052, Japan kanemoto@u-aizu.ac.jpr Health monitoring for rotating machines is investigated through two kinds of mock-up experimental data analysis. One is an anomaly mock-up test of roll bearing type rotating machine. Here, inner ring defect anomaly is simulated and its operating data are measured by both attached type accelerometer sensor and non-attached type microphone. Three kinds of signal pre-processing methods, frequency spectrum, principal component analysis and cepstrum, are applied to discriminate normal and abnormal states using several different classification algorithms, such as adaboost or random forest. Through analysis of their performance with the help of receiver operating characteristic (ROC) curve, the importance of diversified health monitoring methods is discussed. Another mock-up experiment is an accelerated test of roll bearing wear. Here, acoustic emission counts, accelerometer signal and wear particle number in lubricating oil are measured. Using these observation data, we make clear the relationships between deterioration mechanisms of bearing and behaviour of different observations. 1. Introduction Condition based maintenance (CBM) is one of important activities for improving both equipment reliability and maintenance costs. The reliability is expected to increase by avoiding unnecessary overhaul or inspections. In a viewpoint of cost, CBM would contribute not only to reduce maintenance cost but also to allocate limited manpower to other important maintenance work. Reliable health monitoring technology is a key issue to make CBM successful. Based on these general backgrounds, the present paper investigates health monitoring technologies for rotating machines in power plants. To provide reliable health monitoring technologies, it is important to combine various kinds of sensing technologies, signal processing and data classification algorithms. Here, the concept of diversity is important since rotating machines have various kinds of anomaly states. However, we have to take into account that the simple diversity may induce some confusion if diversified monitoring methodologies give different monitoring results. So, it is also important how we can integrate the diversified monitoring results to obtain a reasonable result. Of course, we cannot expect a unique solution for this integration process, but, it is worthwhile to investigate the diversity and integration process by using concrete examples of health monitoring. In the present paper, we will discuss the effectiveness of various kinds of health monitoring methodologies, as well as the importance of integration process of diversified monitoring results, by utilizing two kinds of mock-up experiments of rotating machine anomaly. One is an anomaly mock-up test of roll bearing type rotating machine. Here, normal operating data are compared with inner ring defect data. The other mock- up experiment is an accelerated test of roll bearing wear by adding excessive load on the bearing. In these experiments, various kinds of sensing data were measured, such as accelerometer sensor data, a non-attached type microphone data, acoustic emission data, or, wear particle number in lubricating oil. These data are utilized to investigate what kinds of signal processing and classification algorithms are effective to discriminate normal and abnormal states, or, to predict anomaly progress in the acceleration DOI: 10.3303/CET1333029 Please cite this article as: Kanemoto S., Yokotsuka N., Yusa N., Kawabata M., 2013, Diversity and integration of rotating machine health monitoring methods, Chemical Engineering Transactions, 33, 169-174 DOI: 10.3303/CET1333029 169