Reinforcement Learning as an Approach for Flexible Scheduling

AbstractAfter having defined the contribution of this work in chapter 2, the focus is placed on suitable approaches to solve the extended MrFJSP. As in the case of production scheduling, scheduling problems are often a decision-making process of sequences of situations and decisions within a system of complex relations. It was proven to be efficient to distribute the decision making to independent but cooperating entities, such as the drive agents in the warehouse example.