Digital Twin-Driven AI Framework for Real-Time Monitoring and Autonomous Control of Heavy-Duty Generator Sets with Integrated Bearing Vibration Detection and SCADA Systems

Industrial heavy-duty generator sets are critical assets in modern power generation systems, particularly in oil and gas, manufacturing, utility, and process industries. However, failures associated with bearing degradation, abnormal vibration, thermal instability, and lubrication problems remain major causes of unplanned downtime and operational inefficiency. This paper presents a Digital Twin-driven Artificial Intelligence (AI) framework for real-time monitoring, vibration-based fault diagnostics, predictive maintenance, and autonomous control of heavy-duty gas-powered generator systems integrated with Supervisory Control and Data Acquisition (SCADA) platforms. The framework utilizes operational data obtained from the uploaded Gas Engine Generator (GEG) dataset containing electrical, thermal, mechanical, and combustion-related parameters. The proposed system integrates bearing vibration sensors installed at strategic rotating machine locations for continuous vibration acquisition and intelligent anomaly detection. A hybrid AI architecture combining Long Short-Term Memory (LSTM) networks and Autoencoder models is developed for real-time predictive analytics and fault classification. The Digital Twin continuously mirrors the physical generator behavior and enables intelligent control decisions through SCADA integration. Simulation results demonstrate improved predictive maintenance capability, enhance fault detection accuracy, reduced downtime, and improved operational reliability. The proposed framework provides a scalable solution for Industry 4.0-based intelligent generator monitoring systems.