کنفرانس بین المللی ماشین و محرکه های الکتریکی

| ورود به سامانه | ثبت نام

صفحه اصلی / icemd2024

Acoustic Fault Diagnosis of Rolling Bearings in Induction Motors Using Time-Frequency Image Analysis

نویسندگان :

Erfan Madadi Mahani¹ Ahmad Mirabadi² Amir Hossein Karamali³ Ali Rezazadeh⁴

1- department of Railway Engineering Iran University of Science and Technology 2- department of Railway Engineering Iran University of Science and 3- department of Electrical Engineering Iran University of Science and Technology 4- department of Electrical Engineering Iran University of Science and Technology

کلمات کلیدی :

non-invasive fault detection،condition monitoring

چکیده :

Bearing failure detection in induction motors is critical to avoiding breakdowns and downtime in industrial processes. The analysis of the sound signals and the corresponding spectral signature makes it possible to detect bearing faults in a non-invasive way. For years, several bearing health monitoring systems have been proposed, including the use of sound and vibration sensors. At the same time, analysis of rolling element bearings is a widely accepted practice in the field of condition monitoring of rotating machinery. However, in many cases, the cost of installing an advanced accelerometer- based bearing condition monitoring system, currently the most popular solution in the industry, may be prohibitive due to the potentially long payback period in non-critical machinery. As a result, the human ear is frequently used as the first diagnostic tool, determining the nature of the produced sound. This paper compares machine learning and deep learning approaches to diagnose a rolling element bearing. Then we show how to classify bearing faults by converting 1-D bearing sound signals to 2-D scalogram images and applying transfer learning with a pretrained network. Transfer learning significantly reduces the time required for feature extraction and selection in conventional bearing diagnostic methods, while maintaining good accuracy for the data set used in this work.