Study On Modeling And Optimization Of Key Resource For Muti-product Assembly Plant

Enterprises are facing pressures of early delivery, diversity of customer demandand shorter product life cycles, thus ensuring the delivery time, improving resourceutilization, reducing production costs and improving operational flexibility isimportant for enterprise to keep competitive. This paper will do research on theproblems arising from Daikin Assembly Plant. By solving these problems, productionline efficiency and market responsiveness can be improved, production cycle isshortened. The detail contents are as follows:(1) Modeling of the multi-product assembly lines. Because of the characteristic ofthe air-conditioned assembly line that it is a multi-product assembly line andscheduling is complex, etc, a simulation model of the Assembly Workshop isestablished by using the idea of Object-orient Modeling. This simulation model cannot only be used to verify the production plans and workshop layout, but also providea platform for subsequent verification of the optimal solution.(2) Optimization of material distribution of Picking Area and parallel machines. Asthe cycle time of the assembly line is fast, the speed of picking parts in the pickingarea have certain requirements. We proposed a method to optimize the distribution ofmaterial distribution to shorten the walking distance, thus the speed of picking parts.For Cell Station composed by parallel machines, we’ve proposed four rules toschedule the machines. By comparing the completion time under different rules, thebest rule can be got.(3) Multi-Types AGVs Quantity optimization. To decide the best AGV QuantityConfiguration, we set up the mathematic model using Process、AGV Path、DispatchRule as constraint, making sure the output of workshop largest. Due to thecomplexity of the problem, an optimal method based on Simulation Optimization is proposed, using simulation model to simulate the running of AGVS and the algorithmGA to determine the Quantity of AGVs. The accuracy of optimization result andoptimization speed are effectively improved.(4) Tandem Loop AGV path Optimization. To solve the AGV collision problemcaused by serious crossing of AGV Path, the idea of Tandem Loop Path is proposedto design an AGV path without collision. As Tandem Loop Path exists the problem oftransfer station and flow balance between different Loops, we set up a mathematicmodel to design Tandem Loop Path, using the flow balance between different loops asconstraint and making sure the flow between different loops smallest. For the problemis NP hard problem, a combined algorithm of SA and GA is designed to solve themodel.(5) Application system development. By setting up an simulation model of Daikinassembly plant, and then combining simulation model and solution provided in thispaper, we’ve develop an application system. The feasibility of the solution can beverified by solving the examples based on the problems existing in Daikin assemblyplant.