Digital Twin Platform Based on Virtual Sensing for Real-Time Monitoring of Mechanical Features in Weaving Motion

作者：Pengfei Zhang, Jinrui Zhang, Ruru Pan, Lei Wang, Jian Zhou, Ning Zhang, Jun Xiang · 发表于：IEEE Sensors Journal · 年份：2025 · DOI：10.1109/jsen.2025.3559925 · 被引用次数：3 · 研究领域：Digital Transformation in Industry、Additive Manufacturing Materials and Processes、Engineering Technology and Methodologies

Intelligent monitoring of the weaving workshop, as a key link in the upgrading and transformation of the textile industry, directly affects the production quality and yield of textiles. One of the limitations to efficiently monitoring the weaving workshop is the accuracy and the number of sensors. In this paper, a digital twin model of a weaving workshop was constructed using virtual sensing technology. The virtual weaving workshop, driven by real workshop data, simulated the fabric production process and enabled visual monitoring of the weaving process. According to the motion coordination of the weaving equipment, the simulation of the weaving process was further simplified to meet the timeliness requirements of the actual production. For yarns and fabric, the complex variational operations in mechanics simulations were replaced using a linear elastic model and neural networks. The experimental results demonstrate that the total time, including data acquisition, virtual computation, and visualization, is less than 100 ms. The average accuracy of real-time virtual sensing for yarns and fabrics was 99.8% and 85.1%, respectively. In practice, the optimized digital twin model improves the transfer efficiency of sensing data in web services by 2 times.