基于排队理论的混流制造系统服务率优化与负荷平衡研究

发布时间：2018-04-25 13:45

本文选题：混流制造系统 + 排队理论　；参考：《兰州交通大学》2013年硕士论文

【摘要】：随着市场竞争激烈化以及顾客需求个性化的增加，混流生产方式成为制造业普遍采用的一种生产组织方式，它可以在基本不改变设施布局的前提下，在同一制造系统内组织生产出多种不同型号、不同数量的产品，具有很高的灵活性。但是在多产品混流制造过程中，产品的成组化、产品制造队列的组织、混流制造系统的服务率水平和负荷平衡情况等决定了能否实现制造系统的同步化、均衡化生产，同时，较好的系统服务率水平和均衡稳定的系统生产能力可以使得企业在产品的价格、质量以及快速响应市场能力等方面获得优势，赢得竞争。因此，有必要对混流制造系统的服务率优化问题以及负荷平衡问题进行研究。鉴于排队理论适用于一切服务系统，基于排队理论对混流制造系统进行优化的相关文献并不多见，本文基于排队理论对混流制造系统的服务率优化问题和负荷平衡问题进行研究。针对混流制造系统的高灵活性、产品制造队列排序等特征，利用排队网络理论，可建立一个混流制造系统的开放式排队网络模型。由于整个系统内部存在着多种类型的产品，各类产品可依据成组化原理聚合为一类混合产品，与混合产品相对应的就是将整个混流排队网络模型简化为单一产品类型的排队网络模型，其中的每一个工作单元都可以等效为一个G/G/1排队系统模型。建立了多产品混流制造系统的排队网络模型之后，就可以建立混流制造系统服务率优化问题的数学模型。以每一个工作单元单位时间内产品的逗留损失费用与服务费用之和的期望值为目标函数，以工作单元加工每一类产品的服务强度为约束条件，建立期望费用模型。针对所建立的混流制造系统服务率优化模型的复杂性，构造惩罚函数对原目标函数加以修正，将原约束问题转化为无约束问题。利用坐标轮换法、黄金分割法、进退法等最优化方法以及MATLAB程序对每一个工作单元的惩罚函数进行求解，得到其加工每一类产品的最优服务率，完成混流制造系统的服务率优化工作。确立了每一个工作单元加工每一类产品的最优服务率，即可根据每一类产品在制造系统中的转移概率矩阵，，建立混流制造系统负荷平衡问题的数学模型。根据产品的转移概率矩阵和工作单元的典型队列模型图，利用混流制造系统服务率的优化结果求得每一个工作单元对混合产品的平均加工时间。而后以工作站数目、工作站负荷的平滑指数同时最小为目标函数，以混合产品的加工先后顺序关系为约束条件，建立了混流制造系统负荷平衡的优化模型。运用遗传算法和C程序求解该数学模型，从而完成混流制造系统负荷平衡的工作。最后针对玩具小车混流装配线的服务率优化问题和负荷平衡问题进行实例研究，验证以上优化方案的可行性和实用性。计算结果表明，当混流制造系统每一个工作单元的服务率处于较优状态且系统整体的负荷平衡状况良好时，可以保证生产顺畅，资源利用充分。
[Abstract]:With the intensification of market competition and the increasing personalization of customer demand, mixed flow production has become a kind of production organization widely used in the manufacturing industry. It can produce a variety of different models and different quantities of products in the same manufacturing system without changing the layout of the facilities, but it has high flexibility. In the multi product mixed flow manufacturing process, the group of products, the organization of the product manufacturing queue, the service rate level and the load balance of the mixed flow manufacturing system decide whether the synchronization of the manufacturing system can be realized and the production is balanced. At the same time, the better system service rate and the balanced and stable system production ability can make the enterprise Therefore, it is necessary to study the problem of service rate optimization and load balance in mixed flow manufacturing system.
In view of the fact that queuing theory is applicable to all service systems, there are few related literatures on the optimization of mixed flow manufacturing system based on queuing theory. This paper studies the problem of service rate optimization and load balance of mixed flow manufacturing system based on queuing theory. On the basis of queuing network theory, we can establish an open queuing network model of mixed flow manufacturing system. Because there are many types of products in the whole system, all kinds of products can be aggregated into a class of mixed products according to the principle of grouping, and the whole mixed flow queuing network model is simplified as a single product. Each queuing network model can be equivalent to a G/G/1 queueing system model.
After establishing the queuing network model of multi product mixed flow manufacturing system, the mathematical model of the optimization of the service rate of the mixed flow manufacturing system can be established. The objective function is the expected value of the cost of the lingering loss and the service cost per unit time of each unit, and the service strength of each kind of product is processed by the working unit. According to the complexity of the service rate optimization model of the mixed flow manufacturing system, the penalty function is constructed to modify the original objective function, and the original constraint problem is transformed into unconstrained problem. The optimization methods, such as coordinate rotation, gold segmentation, and retreat method, and MATLAB program to each one are used. The penalty function of each unit is solved to get the optimal service rate of each product, and the service rate optimization of the mixed manufacturing system is completed.
According to the transfer probability matrix of each kind of product in the manufacturing system, the mathematical model of the load balance problem of mixed flow manufacturing system can be established according to the transfer probability matrix of each kind of product in the manufacturing system. The average processing time of each unit to the mixed product is obtained. Then the number of workstations, the minimum of the smoothness index of the workstation load are at the same time as the objective function, and the optimization model of the load balance of the mixed flow manufacturing system is established by the order relation of the processing sequence of the mixed products. The C program solves the mathematical model and completes the load balancing of the mixed flow manufacturing system.
Finally, a case study on the problem of service rate optimization and load balance of the toy car mixed flow assembly line is carried out to verify the feasibility and practicability of the above optimization. The results show that when the service rate of each unit in the mixed flow manufacturing system is in a better state and the overall load balance is good, it can be guaranteed. The production of the certificate is smooth and the resources are fully utilized.

【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TH186

【共引文献】