Developing a digital twin system for a direct-drive electric screw press

Electric screw presses are extensively utilized in industrial manufacturing sectors. However, when operating under extreme conditions, it is prone to performance degradation due to unreliable factors. Additionally, due to limitations in sensor placement and diversity, there are blind spots in monitoring the internal state of the equipment, making it difficult to detect potential faults in a timely manner. To address these challenges, this paper proposes a direct-drive electric screw press (DESP) system integrated with Digital Twin (DT). By accurately simulating the dynamic behavior and performance of physical equipment, DT enables intelligent operation and maintenance functions, including real-time monitoring and fault prediction. This paper presents a comprehensive DT system framework comprising physical, digital, and service layers, with detailed descriptions of implementation methodologies for each module within the digital layer. To validate the proposed approach, a J58ZK-4000 electric screw press was selected as a case study. Using MATLAB Simulink, we built a DT behavior model of DESP and demonstrated its robust capabilities in adaptive updating and data generation. Finally, we briefly outlined sensor data items and showcased the detection of equipment state changes based on AutoEncoder, thereby establishing a foundation for comprehensive state awareness and decision support for DESP.