The Research On Cycle Time Optimization For The Mixed-model Assembly Lines

The mixed-model assembly line is one of the applications of JIT production, because it can meet diversified demand of customers without the need for large product inventories. Cycle time is recognized as a crucial performance parameter of the mixed-model assembly lines, so cycle time optimization becomes a main research field of the mixed-model assembly lines. Scheduling is one of the basic problems in mixed-model assembly lines. The research is taken as follows: Firstly, the thesis analyzes the domestic and foreign research situations on the mixed-model assembly lines, and proposes optimizing the cycle time through solving scheduling problems. The research generalized four modeling methods of the objective function, and selects minimizing the total conveyor stoppage time as the scheduling objective function. Furthermore, the mathematical formulation is presented for the objective function. Secondly, the thesis explains the foundation process of the mixed-model assembly lines optimization model. By comparing the combinatorial optimization algorithms'classes, the thesis recommends the novel ant colony optimization algorithm (ACO) solving this optimization problem. Meanwhile, through the analysis on the basic principle, characteristic, and application of ant colony optimization algorithm (ACO), the thesis puts forward a kind of improvement ACO to establish the mixed-model assembly lines optimization model and to solve the problem and classifications of combinatorial optimization. Thirdly, the thesis elaborates the model's solving process, in which the solution spaces are expressed by graph and matrix after analyzing the mixed-model assembly lines'scheduling feature. Then it searches for the feasible solutions in this solution space, suitable for assembly set's model type and its quantity. According to the model's characteristics, the thesis chooses the reasonable algorithm parameters. Finally, the research gains the satisfied optimal solution by ant colony algorithm's optimal controlling, confirmed by a little example. In the end, by applying Object Oriented technology, this thesis realized the mixed-model assembly lines cycle time optimization system on the Windows'system platform and Visual C++ 6.0 programming environment, verified it via an application instance, and delivered a conclusion and expectation.