Abstract: China is a big textile country. Although the textile industry is an important pillar industry in China and has achieved rapid growth after decades of reform and opening up, backward technology is still an urgent problem to be solved. At present, most of the digital workshops in the textile industry only display the real-time production data of the workshops, and cannot realize the real-time visual monitoring and remote operation and maintenance of the workshops. This kind of workshop has the problems of incomplete information interaction and low degree of visualization, and also produces the “black box” operation, which leads to the insufficient monitoring of the execution information of the underlying workshop by the management personnel. With the emergence of the digital twin technology, the precise real-time mapping of physical entities is realized by constructing a multi-dimensional high-fidelity twin model, which provides an effective solution for the digitization and intelligence of the workshop.
In order to promote the intelligentization and digitization of the traditional wrapping and texturing integrated machine workshop, we study the construction method of the wrapping and texturing integrated machine workshop model based on digital twin. Starting from the digital twin system architecture, we construct the twin model from the four angles of the wrapping and texturing integrated machine, human, yarn and environment. We construct the twin model in three steps. Firstly, we construct the model according to the three dimensions of geometry, physics and production behavior. Secondly, we collect the workshop data. Finally, we map the data accurately to the model in real time. We establish the tree structure chain that binds the geometric relationship and assembly relationship of the model from the geometric dimension, achieve a realistic effect by setting the physical properties and motion state of the object from the physical dimension, and mainly describe the activities of the elastic clad yarn produced by the wrapping and texturing integrated machine from the production behavior dimension. The mapping is mainly divided into position mapping and production behavior mapping. The position mapping changes the object coordinates at the next moment through the coordinate transformation matrix and the object coordinates at the previous moment. The production behavior mapping first defines the correlation matrix according to the processing time and the fact whether the process event is triggered, then determines whether a fault or an abnormal event occurs, and performs integrated analysis according to the classification of event results. Finally, the decision results are fed back to the workshop personnel through the GUI panel and the workshop indicator to achieve the virtual and real linkage. In this paper, the digital twin technology is used to establish the twin workshop model. Compared with the traditional model, it obtains the physical entity data in real time and evolves with the physical entity in real time, so that it is consistent with the physical entity in the whole life cycle of the product. With the wrapping and texturing integrated machine workshop of an enterprise as an example, the effectiveness of the twin model construction method is verified, and the workshop visualization ability, virtual and real interaction ability and production efficiency are improved. 
In this paper, the architecture, construction method and mapping mechanism of the twin model of the workshop are studied, and the virtual simulation of the workshop and the collection and storage of virtual and real data are realized. In the future, the intelligent algorithm will be further studied for fault diagnosis, quality prediction, intelligent decision-making and optimization of production process.

Key words: digital twin, textile industry, virtual-real interaction, wrapping and texturing integrated machine, real-time mapping

摘要： 针对包覆加弹一体机车间虚实交互性差、生产效率低和实时监控难等问题，提出了一种基于数字孪生的包覆加弹一体机车间建模方法。首先搭建包覆加弹一体机车间的数字孪生体系架构，然后从包覆加弹一体机、人、纱线、环境4个角度，阐述孪生模型构建方法，详细介绍了几何属性、物理属性和生产行为；以采集车间生产过程数据为基础，采用Unity 3D和3ds Max 协同开发的方式，对车间物理实体向虚拟车间的位置和生产行为进行实时映射，同时将虚拟车间数据与物理实体数据进行集成分析，优化孪生模型。最后以某企业包覆加弹一体机车间为例，验证了其可行性，有效提高了车间可视化监控能力、车间数据虚实交互能力以及生产效率。

关键词: 数字孪生, 纺织工业, 虚实交互, 包覆加弹一体机车间, 实时映射