Convolutional neural network and 2-D image based fault diagnosis of bearing without retraining

Bearings are vital part of rotary machines. A failure of bearing has a negative impact on schedules, production operation and even human casualties. Therefore, in prior achieving fault diagnosis of bearings is very important. How well features are extracted from vibration signals have a great influence on the performance of traditional intelligent fault diagnosis as well as it is important to achieve good performance without retraining under various operating conditions. However, it usually requires extensive domain expertise and prior knowledge. Instead of traditional machine learning algorithms, deep learning algorithms have a capacity of automatically learning the discriminative feature representation from input data effectively and accurately. So deep learning models can overcome drawbacks of traditional intelligent fault diagnosis. This paper will focus on converting vibration signals to vibration image and then we will use it for convolutional neural network (CNN) which we will use for fault diagnosis to learn features.