Font Size: a A A

Research On Multi-line Parallel Scheduling With Limited Buffers Oriented Agile Final Assembly

Posted on:2014-03-13Degree:MasterType:Thesis
Country:ChinaCandidate:Y J PanFull Text:PDF
GTID:2359330491462792Subject:Mechanical Manufacturing and Automation
Abstract/Summary:PDF Full Text Request
With global market competition increasingly fierce,manufacturing industry is facing great challenges.Rapidly responding to market has been the key to making enterprises competitive,and agile manufacturing mode has become a general trend.In order to meet the requirements of product structure and agile assembly,multi-line with limited buffers production mode appears,in which there is a supply-demand relationship between the upstream and the downstream multi-line.This kind of scheduling problem is more complicated.Empirical research of NFJ multi-line problem,researched on the multi-line parallel scheduling with limited buffers oriented agile final assembly,proved the study has a theoretical value and realistic significance.The main works and achievements can be summarized by the following five points:1.Current research condition and direction of the multi-line scheduling problems is analyzed.Also,the characteristics of the supply-demand relationship between the upstream and the downstream is introduced in the agile environment.Meanwhile,a special resolving idea is put forward which facing to the multi-line parallel scheduling problem with limited buffers.2.Establish the downstream multi-line parallel scheduling model without the constraint of semi-finished products by simulated annealing genetic algorithm and dynamic minimum batch strategy,setting the minimum of the maximum completion time as target function.The new algorithm avoids getting into local optimum and enhances the ability of searching overall optimum.Meanwhile,the key of simulated annealing genetic algorithm is designed.3.According to downstream demands,the upstream multi-line parallel scheduling model is established,which is solved by multi-objective particle swarm optimization algorithm.The objective function of the model is minimized penalty cost and buffer size.And a new particle position update formula is used.4.By mixing programming of Microsoft Visual Basic and MATLAB to realize simulation of multi-line parallel scheduling model with limited buffers oriented agile final assembly,and interface design and program design is done.An example will verify the validity of the scheduling algorithm.5.Empirical research of NFJ multi-line scheduling,simulated the system,obtained the optimal scheduling solution.Contrasted practical scheme and optimal scheduling,optimal scheme reached the better result on completion time and line balance rate of upstream-downstream,proving the feasibility and effectiveness of the model and methed.
Keywords/Search Tags:agile assembly, limited buffers, multi-line parallel scheduling, simulated annealing genetic algorithm, multiple objective particle swarm
PDF Full Text Request
Related items