Font Size: a A A

Reinforcement Learning Based Dynamic Single Machine Scheduling

Posted on:2008-10-04Degree:MasterType:Thesis
Country:ChinaCandidate:S SunFull Text:PDF
GTID:2189360212976469Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
Production scheduling is at the middle layer in the CMIS system structure, which is the joint of control and management. On the one hand, it will supply decisions for the enterprise; on the other hand, it will arrange production tasks and supervise the control layer. So the production scheduling is the key of the CIMS.In the real production, most scheduling problems belong to dynamic scheduling. At present, most dynamic scheduling problems have been addressed using dispatching rules. However, dispatching rules which blindly pursue single goal were basically methods of greedy searching in local area and no dispatching rule can adjust itself according to the change of the status of system's environment and has been found to perform well for all important criteria.Because agent has the ability in adapting itself to dynamic environment and negotiating interactively, many researchers utilize agent technology to solve dynamic problems. But how to improve the learning ability is the key for agent to adapt to dynamic environment changes and...
Keywords/Search Tags:intelligent agent, dynamic single machine scheduling, pattern driven scheduling, reinforcement learning, q-learning algorithm
PDF Full Text Request
Related items