This paper reports a study of correlation dimension in gearbox condition monitoring. In contrast to other fault diagnosis methods, such as Fourier spectrum analysis, timeΒ±frequency analysis, etc., the correlation dimension can provide some intrinsic information of an underlying dynamic system recons
THE USE OF CORRELATION DIMENSION IN CONDITION MONITORING OF SYSTEMS WITH CLEARANCE
β Scribed by C. CRAIG; R.D. NEILSON; J. PENMAN
- Publisher
- Elsevier Science
- Year
- 2000
- Tongue
- English
- Weight
- 193 KB
- Volume
- 231
- Category
- Article
- ISSN
- 0022-460X
No coin nor oath required. For personal study only.
β¦ Synopsis
In this paper the application of embedding space and fractal dimension estimation to monitor the condition of a system with clearance is considered by examining the e!ect on the correlation dimension of variations of the gap clearance in a bearing system. The model analysed represents an elastically supported rotor and stator, subject to excitation by an imbalance, which is restricted to two-dimensional linear movement perpendicular to the axis of rotation. From full-phase space calculations it is found that as the clearance between the rotor and stator is increased, there is a discernible decrease in the value of the correlation dimension. Thus, it may be feasible to determine changing events such as gap clearance by monitoring the correlation dimension of the system. Employing the conventional method of delays results in substandard estimates for the correlation dimension. However, modifying the method of delays by incorporating values from an additional observable into the vectors used for reconstructing the phase space produce results in accordance with those from the full-phase space.
π SIMILAR VOLUMES
Recent works in the vibration-based structural health monitoring community have emphasised the use of correlation dimension as a discriminating statistic in seperating a damaged from undamaged response. This paper explores the utility of attractor dimension as a 'feature' and offers some comparisons
In this paper, some new indices in the sequency domain are defined describing the vibroacoustical process of machinery based on the fast Walsh transform. The Rice frequency is a useful index for detecting the change of power spectral components but it cannot be obtained easily and quickly. Here we d