About this Journal Submit a Manuscript Table of Contents
Advances in Mechanical Engineering
Volume 2013 (2013), Article ID 121340, 2 pages

Prognostics and Maintenance for Mechanical Systems in Harsh Environment

1School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
2Department of Mechanical Engineering, University of Cincinnati, Cincinnati, OH 45221, USA

Received 26 September 2013; Accepted 26 September 2013

Copyright © 2013 Cheng-liang Liu et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1. Introduction

The idea of preparing this special issue aims at solving emerging problems when condition monitoring is applied to more and more machines running in practical harsh environments. Such machines, including wind turbines, mining and civil engineering machinery, are not easy to be frequently checked and maintained by man power. Most established machine condition monitoring are developed in ideal lab-setting conditions. Considering the real-world highly dynamical load and harsh working conditions, there is a need to develop practical ways to handle various situations and rapidly response to the faults and significant degradation process. Thus, the authors of this special issue were asked to prepare papers on the following topics:(i)diagnosis and maintenance of engineering equipment within extreme environment,(ii)mechanical structure analysis under extreme environment,(iii)decision-making strategies for reliable operations of mechanical systems,(iv)information/data mining on mechanical system maintenance,(v)robust, self-sustainable remote sensing, and monitoring systems.

2. Prognostics and Maintenance for Mechanical Systems in Harsh Environment: Past, Present, and Future

The safety, reliability, and maintainability of mechanical equipment during their service life are issues of critical importance to be solved by either the manufacturing industry or the service suppliers. In the past, for the machines operating in harsh environment, the reliable and robust design was the only aspect that people could do their best to keep them in the reliable service time as long as possible. The traditional “fail and fix (FAF)” practice was adopted at that time. However, unexpected downtime often may cause great loss because the maintenance could take so much time to repair, or even the replacement of the faulty machine with a new one becomes a must-do due to severe chain damages. Therefore, firstly, people began to assume some certain level of performance degradation for the machinery, without input from the working load or any sensors with itself, and serviced equipment on a routine schedule, whether the service is actually needed or not, which may be called blindly proactive maintenance. With fast development of sensor and tether-free communication technologies, condition-based maintenance (CBM) is found to be more effective approaches in achieving the goals of reliability and maintainability for this mechanical system running in harsh environment, which has become a value-added service, apart from the quality of products in design and manufacturing. The CBM also makes it possible to measure the faults more precisely and based on the evidence on data from running machine, severe unexpected faults can be avoided by scheduling the pre-cautious maintenance before they actually occur. Thus, the prognostics shifts the maintenance into the “predict and prevent (PAP)” paradigm. From then on, more and more challenges emerge with wider and wider applications of prognostics, for example, the disturbances in the monitored signals under working process, the problem of feature extraction from long sampling signals, and some inherent disadvantages in the feature-based machine condition recognition. In order to effectively extract the information carried by the monitoring data, redundancy reduction techniques may be taken. The feature-based machine condition recognition also has some inherent disadvantages, such as the “inevitable misdiagnosis” problem and “lacking of historical monitoring data”. For the problem of how to overcome the disturbances under the working process, the blind separation technique based on sparse component analysis could be applied. Other new trends for better applying prognostics may include the methods related to IoT (Internet of Things), cloud computing, big data, and the refactoring database methodology. With data mining technology, the intelligence models should be formed as generative models that can self-detect and self-heal the mechanical systems.

3. Special Issue Overview

In this special issue, the above-mentioned trends are embodied by pioneer research. Classical digital signal processing methodologies, such as Wavelet and EMD, with improved performance and targeted application by W. Du et al. in “Wavelet leaders based vibration signals multifractal features of plunger pump in truck crane” and Q. Yang et al. in “EMD and Wavelet transform based fault diagnosis for wind turbine gear box” are presented to rapidly and precisely mine the fault data. One of the most challenging issues in prognostics and maintenance fields is to manipulate the incomplete data for evaluating the mechanical systems’ performance. In the paper titled “Study on immune relevant vector machine based intelligent fault detection and diagnosis algorithm,” Z. Miao et al. proposes a “self/nonself” identification principle to effectively tackle this problem with the inspiration of artificial immune system. To efficiently extract features from big data and conduct the dimension-reduction processing, a condition-based monitoring system that is tailored to the embedded deployment is demonstrated in “A computationally efficient and adaptive approach for online embedded machinery diagnosis in harsh environments” by C. Jiang and S. H. Huang Boosted by the viewpoint of the whole system performance assurance, in “Performance assessment for a fleet of machines using a combined method of ant-based clustering and cmac,” L. Zhang et al. advocate this concept and the relevant analyzing techniques. As spin-off hot topics, some cross-discipline analysis for electrical mechanical system performance is also embraced in this issue, such as “Condition evaluation of large generator stator insulation based on partial discharge measurement,” which highlights the fault detection and quantitative assessment procedures in harsh environment.

Cheng-liang Liu
Jay Lee
Liang Gong
Yixiang Huang