Condition monitoring

About: Condition monitoring is a research topic. Over the lifetime, 13911 publications have been published within this topic receiving 201649 citations.

A hybrid feature selection scheme for unsupervised learning and its application in bearing fault diagnosis

Abstract: With the development of the condition-based maintenance techniques and the consequent requirement for good machine learning methods, new challenges arise in unsupervised learning. In the real-world situations, due to the relevant features that could exhibit the real machine condition are often unknown as priori, condition monitoring systems based on unimportant features, e.g. noise, might suffer high false-alarm rates, especially when the characteristics of failures are costly or difficult to learn. Therefore, it is important to select the most representative features for unsupervised learning in fault diagnostics. In this paper, a hybrid feature selection scheme (HFS) for unsupervised learning is proposed to improve the robustness and the accuracy of fault diagnostics. It provides a general framework of the feature selection based on significance evaluation and similarity measurement with respect to the multiple clustering solutions. The effectiveness of the proposed HFS method is demonstrated by a bearing fault diagnostics application and comparison with other features selection methods.

Evaluation of expert system for condition monitoring of a single point cutting tool using principle component analysis and decision tree algorithm

Abstract: Tool wear and tool life are the principle areas are focus in any machining activity. The production rate, surface finish of machined component and the machine condition are directly related to the tool condition. This work on tool condition monitoring delves into data mining approach to discover the hidden information available in the tool vibration signals. The use of statistical features derived from the vibration data is used as the primary feature and Principle Component Analysis (PCA) transformed statistical features are evaluated as an alternative. In order to increase the robustness of the classifier and to reduce the data processing load, feature reduction is necessary. The feature reduction using (a) decision tree and (b) feature transformation and reduction using PCA are evaluated independently and the results are compared. The effective combination of feature reducer and classifier for designing the expert system is studied and reported.

Application of UHF Sensors in Power System Equipment for Partial Discharge Detection: A Review

Abstract: Condition monitoring of an operating apparatus is essential for lifespan assessment and maintenance planning in a power system. Electrical insulation is a critical aspect to be monitored, since it is susceptible to failure under high electrical stress. To avoid unexpected breakdowns, the level of partial discharge (PD) activity should be continuously monitored because PD occurrence can accelerate the aging process of insulation in high voltage equipment and result in catastrophic failure if the associated defects are not treated at an early stage. For on-site PD detection, the ultra-high frequency (UHF) method was employed in the field and showed its effectiveness as a detection technique. The main advantage of the UHF method is its immunity to external electromagnetic interference with a high signal-to-noise ratio, which is necessary for on-site monitoring. Considering the detection process, sensors play a critical role in capturing signals from PD sources and transmitting them onto the measurement system. In this paper, UHF sensors applied in PD detection were comprehensively reviewed. In particular, for power transformers, the effects of the physical structure on UHF signals and practical applications of UHF sensors including PD localization techniques were discussed. The aim of this review was to present state-of-the-art UHF sensors in PD detection and facilitate future improvements in the UHF method.

Wavelet transform with spectral post-processing for enhanced feature extraction

Abstract: The quality of machine condition monitoring techniques as well as their applicability in the industry are determined by the effectiveness and efficiency with which characteristic signal features are extracted and identified. Because of the weak amplitude and short duration of structural defect signals at the incipient stage, it is generally difficult to extract hidden features from the data measured using conventional spectral techniques. A new approach based on a combined wavelet and Fourier transformations is presented in this paper. Experimental studies on a rolling bearing with a localized point defect of 0.25 mm diameter has shown that this new technique provides significantly improved feature extraction capability over the spectral techniques.