Dissertation
Dissertation > Economic > Economic planning and management > Economic calculation, economic and mathematical methods > Economic and mathematical methods

Multi-agent -based flexible job shop pre / re- scheduling system

Author ZouYanRui
Tutor RenHaiYing
School Beijing University of Technology
Course Management Science and Engineering
Keywords Pre / re- scheduling Multi-agent Flexible job-shop Contract Net Protocol
CLC F224
Type Master's thesis
Year 2011
Downloads 94
Quotes 0
Download Dissertation

Scheduling is how some tasks can be arranged on the machine to achieve optimal objective function of the problem. The complexity of scheduling, dispatching and scheduling uncertainty of the environment so that the dynamic nature of the task scheduling is becoming increasingly difficult, in order to better deal with this situation, the scheduling system must have strong robustness, scalability, and Adaptability. Flexible job shop scheduling problem to reduce machine constraints, increasing the complexity of the problem, closer to the actual working environment, has important theoretical value and engineering significance. Pre / Rescheduling is used to solve the problem of dynamic scheduling method, which combines the complete reaction robust scheduling method for scheduling and the advantages and overcome their shortcomings, both a dynamic and fast response capability in the external environment and has certain Lu Stick sexual characteristics. Rescheduling workpiece affected thinking but also greatly reduces the implementation process of rescheduling the conversion costs, and better achieve system optimization and ensure the stability of the system. Multi-agent system, the basic idea is to decompose the complex system, to solve the problem through negotiation mechanism. Because of its dynamic flexibility and adaptability, production scheduling system is increasingly becoming an important research methods. Based on the research and multi-agent theory based on multi-agent system is designed based on multi-agent flexible job shop scheduling system, the Scheduling System workpiece physical entities, processes and machines mapped to agent, through consultations between the scheduling agent programs. In this sense, negotiation strategies for scheduling system performance has a pivotal position, this is the traditional basis of contract net protocol has been extended to support many-negotiation mechanism for more effective collaboration between the completion of the agent. In this paper, discrete event simulation method, based on the java language in eclipse platform to achieve scheduling system simulation. Multi-target strategy to an average lag as the main objective, the total processing time, mean flow time and machine utilization as secondary objectives scheduling scheme for multi-angle analysis. First produced by pre-negotiation strategy in advance scheduling scheduling scheme, and with the Jain genetic algorithm, which demonstrates the effectiveness of multi-agent approach. On the basis of pre-scheduling scheme, considering the three types of dynamic scheduling environment interference, machine failure, urgent arrival and workpiece cancel interference were designed based on agent handling strategy and objectives and begin a pre-scheduled departure time for the target simulation experiments with conventional right-shift method, which demonstrates excellent resistance method. The proposed multi-agent-based flexible job shop pre / re-scheduling system to better meet the dynamic shop scheduling system, integration, smart and practical requirements.

Related Dissertations
More Dissertations