Stochastic parallel machine scheduling using reinforcement learning

Abstract

In a high‐mix and low‐volume manufacturing facility, heterogeneous jobs introduce frequent reconfiguration of machines which increases the chance of unplanned machine breakdowns. As machines are often nonidentical and their performance degrades over time, it is critical to consider the heterogeneity and non‐stationarity of the machines during scheduling. We propose a reinforcement learning‐based framework with a novel sampling method to train the agent to schedule heterogeneous jobs on non‐stationary unreliable parallel machines to minimize weighted tardiness. The results indicate that the new sampling approach expedites the learning process and the resulting policy significantly outperforms static dispatching rules.

Deep Reinforcement Learning-Based Scheduler on Parallel Dedicated Machine Scheduling Problem towards Minimizing Total Tardiness

“…More recently, deep Q-network-based schedulers were suggested for semiconductor manufacturing applications [21,22]. A deep deterministic policy gradient (DDPG)-based scheduler was also proposed to minimize weighted tardiness in the stochastic parallel machine scheduling problem [23]. Unlike [19,20], these studies developed multi-agent approaches where each agent considers the allocation of a job on a machine, and they successfully improved performances by reducing the learning complexity.…”