A New Joint Data-Model Driven Dynamic Scheduling Architecture for Intelligent Workshop

Abstract

Abstract Intelligent manufacturing plays a significant role in Industry 4.0. Dynamic shop scheduling is a key problem and hot research topic in the intelligent manufacturing systems, which is NP-hard. However, traditional shop scheduling mode, dynamic event prediction approach, scheduling model and scheduling algorithm, cannot cope with increasingly complicated problems under kinds of scales production disruptions in the real-world production. To deal with these problems, this paper proposes a new joint data-model driven dynamic scheduling architecture for intelligent workshop. The architecture includes four new and key characteristics in the aspects of scheduling mode, dynamic event prediction, scheduling model and algorithm. More specifically, the new scheduling mode introduces data analytics methods to quickly and accurately deal with the dynamic events encountered in the production process. The new prediction model improves the deep learning method, and further applies it predict the dynamic events accurately to provide reliable input to the dynamic scheduling. The new scheduling model proposes a new hybrid rescheduling and inverse scheduling model, which can cope with almost scales of abnormal production problems. The new scheduling algorithm hybridizes dynamic programming and intelligent optimization algorithm, which can overcome the disadvantages of the two methods based on the analysis of solution space. The dynamic programming is employed to provide high-quality initial solutions for the intelligent optimization algorithm by reducing the computation time greatly. To sum up, the presented architecture is a new attempt to understand the problem domain knowledge and broaden the solving idea, which can also provide new theories and technologies to manufacturing system optimization and promote the applications of the theoretical results.