ABSTRACT

One of the significant trends in smart manufacturing is the idea of industrial digitalization, which is enabled through the use of new information technologies, such as the Internet of Things, big data, cloud computing, and artificial intelligence. However, manufacturing industries can only be achieved by combining the physical manufacturing world and digital world, to realize a series of smart manufacturing activities, such as active perception, real-time interaction, automatic processing, intelligent control, and real-time optimization, etc. In this paper, a digital twin-driven approach combines with agent-based decision-making for real-time optimization of motion planning in robotic cellular is proposed, with optimizing the physical and virtual layer at the manufacturing facility. Accordingly, an architecture of the digital twin-driven facility is design, and its operational mechanisms and implementation methods are explained in detail. Moreover, qualitative analysis and a quantitative comparison based on a real robotic cell are provided. Several key findings and observations are generated relating to managerial implications, which are valuable for various users to make manufacturing decisions under the digital twin-driven environment.

摘要

智能制造的重要趋势之一是工业数字化的理念，它通过使用新的信息技术，如物联网、大数据、云计算和人工智能来实现。然而，制造业只能通过物理制造世界和数字世界相结合，实现主动感知、实时交互、自动加工、智能控制、实时优化等一系列智能制造活动。在本文中，提出了一种数字孪生驱动方法与基于代理的决策相结合，用于实时优化机器人细胞中的运动规划，并优化制造设施的物理层和虚拟层。因此，设计了数字孪生驱动设施的架构，详细阐述了其运行机制和实现方法。此外，还提供了基于真实机器人单元的定性分析和定量比较。产生了与管理影响相关的几个关键发现和观察结果，这对于不同用户在数字孪生驱动的环境下做出制造决策很有价值。