Development and Utilization of Synthetic Signals for Fault Diagnostics of Electrical Machines

Abstract

The industrial revolution has opened up more paths with the integration of information technology with industrial applications. Similarly, most industrial processes can be streamlined by combining the Internet of Things and artificial intelligence. Artificial intelligence has a significant role in this development, whether it is related to real-time condition monitoring of electrical machines or switching of the industry from scheduled maintenance to predictive maintenance. One of the main challenges for artificial intelligence is the quality and quantity of data used for training models, as it requires big datasets to train more accurate and efficient models. This article presents a data acquisition system with real-time condition monitoring of electrical machines. A comparison between trained models from real signals and synthetic signals, generated through the equation, is also covered in this article. This is to help identify whether utilizing synthetic signals for the training of fault diagnostics models can be a good alternative in the long run or not.