Electrical machines: Real-time Simulation for Intelligent Fault Diagnosis, Prognostics and Health Management

This article reviews selected research focused on digital real-time simulation for the intelligent condition monitoring of electrical machines. Traditional multi-physical digital simulations have been broadly used in industrial product manufacturing and processing, particularly in virtual testing under regular and harsh working conditions, aging, and life-cycle analysis. Recent progress in data communication and information technologies based on virtual reality, cloud computing, parallel processing, machine learning, large-scale/big data, and the Internet of Things has allowed the development of intelligent physical and mathematical models with the primary capability of real-time bilateral data transfer with a physical unit. This new trend leads to perform condition-based monitoring, fault diagnosis, prognostics and life-cycle analysis more efficiently. An overview of the recent progress and challenges in the above research topic is presented. The primary emphasis is on the real-time multi-physical modeling of electrical machines, which are broadly utilized in industrial plants.