A hybrid model for bearing performance degradation assessment based on support vector data description and fuzzy c-means

Abstract

Bearing performance degradation assessment is more effective than fault diagnosis to realize condition-based maintenance. In this article, a hybrid model is proposed for it based on a support vector data description (SVDD) and fuzzy c-means (FCM). SVDD, which holds excellent robustness to outliers, is used to obtain the clustering centre of normal state. The subjection of tested data to normal state is defined as a degradation indicator, which is computed by a FCM algorithm with final failure data. The results of applying this hybrid model to an accelerated bearing life test show that it can effectively assess bearing performance degradation. Furthermore, it is robust to the outliers in the training set and is not influenced by the Gaussian kernel parameter.

Analysis of load rejection transient process based on static efficiency factor for hydraulic unit security assessment

“…11 For this issue, extensive studies have been done theoretically, experimentally and numerically, and useful measures have been demonstrated. 12,13 For example, Warda and Kandil used the simple method of characteristics (MOC) to monitor pressure transients in pipes. It is concerned with finding a universal model for unsteady friction in both laminar and turbulent flows.…”

Prognostics of slurry pumps based on a moving-average wear degradation index and a general sequential Monte Carlo method

“…Liao and Lee [18] proposed a novel degradation assessment method based on data collected from transient periods of different working loads. Pan et al [19] combined wavelet packet transform and a fuzzy c-means to assess bearing health condition and then developed a hybrid method [20], that consists of a support vector data description and a fuzzy c-means, to evaluate bearing health condition. Wang et al [21] used a series of wavelet filters to extract gear fault features and employed a support vector data description to track the current health condition of a gear.…”

Nonlinear prediction of condition parameter degradation trend for hydropower unit based on radial basis function interpolation and wavelet transform

“…Performance degradation assessment of equipment is a new expansion for the existing research field. [1][2][3][4][5][6] It is more focused on the performance degradation process in the entire life cycle of equipment, and then can comprehensively assess the operational condition of equipment. In the process of performance degradation for equipments, if the degradation degree of equipments performance can be identified, it can be targeted to organize the production and maintenance plan of equipment, prevent equipment malfunction and failure.…”