Adaptive Scheduling of Robots in the Mixed Flow Workshop of Industrial Internet of Things

With the deep integration of industrial Internet of Things technology and artificial intelligence technology, the material robot has been widely used in the Internet of Things workshop. In view of many complex factors such as real-time dynamic change and uncertain condition in workshop, this paper proposes to realize workshop adaptive scheduling decision with component layer construction and SPMCTS search method with real-time state as the root node. This method transforms the robot scheduling problem into a Markov decision process and describes a detailed representation of workshop states, actions, rewards, and strategies. In the real-time scheduling process, the search method is based on the artifact component layer construction, and only considers the state relationship between two adjacent groups, so as to simplify the calculation difficulty. In the subtree search, SPMCTS is applied to search the real-time state as the root node, and the extension method and shear method are applied to conduct strategy exploration and information accumulation, so that the deeper the real-time state node in the subtree, the more the optimal strategy can be obtained quickly and accurately. Finally, the effectiveness and superiority of the proposed method are verified by real case simulation analysis.