Akademik Veri Yönetim Sistemi
INTERNATIONAL CONGRESS ON NATURAL AND ENGINEERING SCIENCES, Siirt, Türkiye, 22 - 24 Kasım 2024, ss.22, (Özet Bildiri)
Bearing failures are a common challenge in mechanical systems, significantly affecting performance and operational reliability. Also rolling bearings are one of the mechanical parts that have been studied the most because of their extensive usage in industrial systems and the need of predictive maintenance. Accurate prediction of these failures is critical for effective maintenance planning and preventing unexpected breakdowns. Implementing predictive maintenance can revolutionize industrial equipment upkeep by minimizing unplanned downtime and mechanical failures, leading to substantial savings in labor, materials, and costs. Predictive maintenance is not yet widely used in industry, but research in this field will eventually result in its broad use. Additionally, predictive maintenance will reduce the loss of labor, material resources, and expenses caused by unplanned malfunctions. Because of these predicting the remaining life of the bearing is essential for predictive maintenance. The remaining life of a bearing can be estimated by processing its vibration data. Converting the vibration signal to an image and using CNN models is one of the finest ways to handle it. Bearing vibration signals can be transformed by an image Wavelet Transform and the remaining life of the bearing is predicted with a CNN like ResNet50 CNN model.