.png){kind=icon}
.png){kind=icon}
Return
Minimizing Model Complexity in Fault Classification Using Inter-Axis Phase Difference
.png){kind=icon}
PRE
AI
DOI：10.1007/978-3-032-04160-9_9.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Predictive maintenance, which aims to detect faults early and perform the necessary maintenance, is important and has been extensively researched. In fault detection using vibration data, phase information is very useful for early detection. For this study, we focus on methods that use both amplitude and phase information for fault diagnosis. To effectively use phase information, preprocessing steps such as background noise removal are required. Therefore, focusing on the characteristics of vibration data is considered essential. We propose a phase correction method that preserves inter-axis phase information and a phase restriction method based on a reference frequency. As the reference frequency, the rotational frequency of the machinery is adopted. In the inter-axis synchronization correction method, preserving the relative phase information between axes in multi-axis vibration data is expected to improve classification accuracy. The phase restriction method aims to enhance both accuracy and model efficiency by reducing information that may prevent classification. As a result, the former contributed to improved accuracy, and the latter achieved simplification of the model.
Keywords:
Predictive maintenance
Deep learning
Inter-axis syncronize
Journal
D
IF:
0
Papers:
34
Citations:
0