Sequencing with earliness and tardiness penalties: a review
Operations Research
Neuro-Dynamic Programming
Multi-Machine Scheduling - A Multi-Agent Learning Approach
ICMAS '98 Proceedings of the 3rd International Conference on Multi Agent Systems
Learning to act using real-time dynamic programming
Artificial Intelligence
Hi-index | 0.00 |
A new distributed learning and control (DLC) approach is presented in this paper which integrates part-driven distributed arrival time control (DATC) and machine-driven distributed reinforcement learning control. This approach is suitable for just-in-time (JIT) production for multi-objective scheduling problem in dynamically changing shop floor environment. While part controllers are adjusting their associated parts' arrival time to minimize due-date deviation, machine controllers equipped with learning components are searching for optimal dispatching policies. The machines' control problem is modeled as Semi Markov Decision Process (SMDP) and solved using one-step Q-learning. The DLC algorithms are evaluated for minimizing the sum of duedate deviation cost and setup cost using simulation. Results show that DLC algorithms achieve significant performance improvement over usual dispatching rules in complex real-time shop floor control problems for JIT production.