Journal of Electrical Engineering and Automation

The accuracy of digital twin models in the process industry directly affects their application effects in real-time optimization and fault prediction. This paper focuses on the research of "real-time calibration methods for digital twin models in process industry" and proposes a dynamic calibration strategy based on the integration of data-driven and mechanism-based approaches. The strategy aims to enhance the model's adaptability to complex working condition changes. By constructing a multi-source data acquisition framework and an online parameter update mechanism, the continuous approximation between model output and the actual system state is achieved. Experimental results show that this method can significantly improve the real-time performance and accuracy of the model, providing strong support for the intelligentization of the process industry.

Digital twin; Real-time calibration; Process industry; Data-driven; Model optimization