In recent years, the manufacturing industry has faced various global challenges. One of these challenges is the increasing frequent adjustment and reconfiguration of production lines, necessitated by the diversification of customer demands. To improve the processing efficiency after production line reconfiguration, this paper puts forward a group learning architecture towards intelligent equipment. With this architecture, swarm intelligence can be accomplished through group learning. From the perspective of edge intelligence, this paper addresses key technology issues within the areas of data acquisition and preprocessing, cyber-physical fusion, knowledge extraction and sharing, and equipment performance self-optimization. The proposed approaches are much more useful for improving the processing efficiency of the reconfigured production line.

近年来，制造业面临着各种全球性挑战。这些挑战之一是，由于客户需求的多样化，生产线的调整和重新配置变得越来越频繁。为了提高生产线重构后的处理效率，本文提出了面向智能设备的小组学习架构。使用这种架构，可以通过小组学习来实现群体智能。从边缘智能的角度，本文解决了数据采集和预处理，网络物理融合，知识提取和共享以及设备性能自我优化等领域中的关键技术问题。所提出的方法对于提高重新配置的生产线的处理效率非常有用。