Research On Simulation And Optimization Of Dynamic Job Scheduling In Re-entrant Manufacturing System

Re-entrant manufacturing system is a new type of manufacturing mode that accompanies the development of the semiconductor industry,and has gradually became popular with modern machining centers and information technology.Therefore,it is of great significance to study the job scheduling strategy of re-entrant manufacturing system to improve production efficiency,improve service levels,and reduce production costs.This dissertation chose the re-entrant manufacturing system as the research object and the dynamic job scheduling as the research perspective,established a simulation platform that integrated the priority scheduling rule and the DBR(Drum-Buffer-Rope)dynamic scheduling strategy by system simulation method.Through this platform,a dynamic job scheduling strategy was optimized for an actual case.The details are as follows:Firstly,this dissertation studied the modeling,grading,combination,and evaluation methods of priority scheduling rules.This dissertation established mathematical models for the seven types of priority scheduling rules that can be used in the simulation platform.Combined with the re-entrant characteristics of re-entrant manufacturing system,the multi-priority scheduling rule which considering re-entrant factors and the combined priority scheduling rule which different process call scheduling rules independently are studied.As well as combining the multi-objective characteristics of re-entrant manufacturing system,the TOPSIS multi-objective planning method was introduced to evaluate priority scheduling rule.Secondly,this dissertation studied the theory and implementation of DBR scheduling strategy.This dissertation presented four kinds of key bottleneck resource identification methods,and established the range of critical bottleneck buffers by the “minimum maximum process time method”.Besides,two kinds of DBR scheduling policy control strategies(batch buffering strategy and interval buffering strategy)and their parameter configuration method were proposed.It also introduced two possible risks(bottleneck drift and bottleneck hunger)that may occur in the DBR scheduling process and their prevention methods.Then,this dissertation established a platform framework for the job scheduling simulation platform for re-entrant manufacturing system.Based on this framework,a simulation platform which integrated priority scheduling rule?multi-level priority scheduling rule?combined priority scheduling rule and DBR dynamic scheduling strategy was built by Plant Simulation software and modular modeling methods.Besides,the simulation platform contains TOPSIS multi-objective planning methods and genetic algorithm,so that it can evaluate and optimize re-entrant manufacturing system quickly and effectively.At the same time,a GUI user interaction interface was established for the simulation platform,which reduced the user's usage threshold.Finally,this dissertation built a model for an example through the dynamic job scheduling simulation platform for re-entrant manufacturing system.Based on this model,this dissertation optimized this example from the type of priority scheduling rule?multi-level priority scheduling rule?combined priority scheduling rule and DBR dynamic scheduling strategy.Besides,the simulation platform got the parameters of dynamic interval buffer strategy and dynamic batch buffer strategy by genetic algorithm.Lastly,the optimal dynamic job scheduling combination strategy was found by genetic algorithm.