A hybrid approach for improving the flexibility of production scheduling in flat steel industry

Abstract

Nowadays the steel market is becoming ever more competitive for European steelworks, especially as far as flat steel products are concerned. As such competition determines the price products, profit can be increased only by lowering production and commercial costs. Production yield can be significantly increased through an appropriate scheduling of the semi-manufactured products among the available sub-processes, to ensure that customers’ orders are timely completed, resources are optimally exploited, and delays are minimized. Therefore, an ever-increasing attention is paid toward production optimization through efficient scheduling strategies in the scientific and industrial communities. This paper proposes a hybrid approach to improve the flexibility of production scheduling in steelworks producing flat steel products. Such approach combines three methods holding different scopes and modelling different aspects: an auction-based multi-agent system is applied to face production uncertainties, multi-objective mixed-integer linear programming is used for global optimal scheduling of resources under steady conditions, while a continuous flow model copes with long-term production scheduling. According to the obtained simulation results, the integration and combination of these three approaches allow scheduling production in a flexible way by providing the capability to adapt to different production conditions.

3D reconstruction based on hierarchical reinforcement learning with transferability

“…The combination of transfer learning (TL) and RL has attracted considerable attention from researchers in order to improve the efficiency of RL, such as policy distillation [63,64], and learning from demonstrations [65,66], and so on. The combination of various methods can compensate for each other's shortcomings [67,68]. However, the knowledge of complex agents can easily fail to be transferred since the inconsistent state distribution among data of different classes, worse still may result in negative impact.…”

Enhanced memetic search for reducing energy consumption in fuzzy flexible job shops

“…Scheduling [1,2] is a classical manufacturing problem that consists in organizing the execution of a set of tasks or operations to make the best use of the available resources under some constraints in order to fulfill some objectives. Solving scheduling problems is essential not only in industrial applications [3,4], in particular in Industry 4.0 and 5.0 [5], but also in healthcare [6,7], computing infrastructures [8] or education [9]. Besides its undeniable applicability, scheduling also poses a computational challenge, since many of its problems are NP-Hard [10].…”