Edge Computing: A Promising Framework for Real-Time Fault Diagnosis and Dynamic Control of Rotating Machines Using Multi-Sensor Data

Edge computing is an emerging computing paradigm where computation is executed on the edge of networks rather than on cloud servers, thereby reducing the system response time, transmission bandwidth occupation, and storage and computation resources on the cloud. In this paper, an edge computing-based method for real-time fault diagnosis and dynamic control of rotating machines is proposed, and an edge computing node (ECN) is designed. A vibration signal and three motor-phase current signals are acquired synchronously by the ECN. Subsequently, 40 features are extracted from the signals and fused into an indicator that can effectively distinguish multiple motor conditions. The experimental results show that the designed ECN can diagnose electrical and mechanical faults with 100% accuracy and take control of the motor when an emergency fault is detected. The algorithm procedures, including signal acquisition, feature extraction and fusion, and fault identification, can be accomplished within 0.25 s on the ECN. The proposed edge computing framework processes sensor data in real time and thus shows potential applications in the rotating machines where fault diagnosis and dynamic control are highly time sensitive.