基于连续批加工的生产运输协同调度研究

发布时间：2018-05-12 10:10

本文选题：调度 + 连续批　；参考：《合肥工业大学》2014年博士论文

【摘要】：随着网络技术和全球经济一体化的不断发展,供应链之间的竞争越来越激烈。更多的供应链成员意识到需要通过加强与其他成员之间的合作来提高供应链的竞争力,从而能够降低各自的生产运营成本。物联网技术的发展正好为供应链成员之间的合作提供了信息基础,它不仅能够将生产和运输信息实时反馈到各自成员的管理中心,而且能够及时将这些信息分享给其他合作成员。物联网技术将供应链成员之间的合作推进到一个新的层面,利用好这些信息能够低生产成本、增加利润和提高客户满意度,从而增强整体供应链竞争力,同时能够进一步拓宽了生产运输协同调度问题的理论研究领域。因此,如何将该信息价值转化为经济和社会效益,运用物联网信息获得高效的生产运输协同方案成为关键性问题。本文以铝制品制造供应链为背景,从调度角度出发研究了物联网环境下供应链成员之间的生产运输协同调度问题。本文围绕挤压厂的连续批处理机加工过程,系统性分析了多种情形下的多阶段生产运输协同调度问题。分别考虑了运载车辆有限情形、工件动态到达情形、工件加工时间恶化情形、机器发生故障情形以及分布在不同地理位置的多制造商情形的协同调度问题。由于这些问题均为NP难,本文致力于分析最优调度方案的性质,并基于这些性质设计高效的启发式和智能算法。另一方面,本文为这些问题推导了问题下界,这些下界可以用于评价算法的精确度。本文对基于连续批的生产运输协同调度问题的研究成果概括如下： (1)研究了运载车辆有限情形下的生产运输协同调度问题,调度目标为最小化制造跨度时间。基于运载车辆有限等约束条件,建立了数学模型。根据批次在供应商和制造商之间的运输时间和工件在制造商机器上加工时间的关系,将问题划分为两种情形。针对两种情形,分别设计了两种启发式算法及问题下界。基于问题下界推导出该启发式算法最坏情形性能比例。考虑工件数量规模,分别设计了大量的小规模和大规模随机数据实验。仿真实验结果表明,当工件数量为1000时,所构造的两种启发式算法的平均相对差距分别收敛到0.48%和0.80%,均明显优于现有FOE and SPT算法和LOE and LPT算法。 (2)研究了工件动态到达情形下的生产运输协同调度问题,调度目标为最小化制造跨度时间。基于工件动态到达等约束条件,建立了数学模型。分析了最优调度方案中关于工件动态到达时间的相关性质,讨论了工件同时到达特殊情形最优调度方案和批次数量之间的关系。基于以上问题性质,构造了两阶段TP-H启发式算法,并证明TP-H算法最坏情形性能比例为7/2。分别通过松弛工件到达时间和假设制造商机器上没有额外空闲时间得出两个问题下界。基于不同的机器能力设计了仿真实验,大量仿真实验结果验证了TP-H算法的有效性。当工件数量为1000时,.所构造的TP-H算法的平均相对差距收敛到0.21%,优于现有MBF算法和MEF算法。 (3)研究了工件加工时间恶化的生产运输协同调度问题。分析了工件加工恶化情形下的几类单机调度问题,目标函数分别为最小化制造跨度时间、最小化延迟工件数量、最小化工件总完工时间,设计了相应的最优化算法。基于单机调度问题,考虑了有缓存区情形下的生产运输协同调度问题,建立了相应的数学模型,目标函数为最小化制造跨度时间。分析了最优调度方案中组批及批次排序性质,构建了一种最优化算法。另外研究了无缓存区情形下的生产运输协同调度问题,建立了基于无缓存区约束的数学模型,目标函数为最小化制造跨度时间。分析了最优解性质,推导出新的问题下界,构建了一种启发式算法。仿真实验结果表明当工件数量大于260时,该启发式算法的平均和最大相对差距比例均小于0.01%。 (4)研究了考虑机器故障的生产运输协同调度问题。基于两个平行制造商机器可能发生故障的情形建立了数学模型,目标函数为最小化制造跨度时间。分析了最优调度方案性质,推导出问题下界。分别构建了两种情形下的调度规则,基于这些调度规则,设计了新的启发式算法。大量的仿真实验结果说明了该启发式算法能够高效地解决不同规模的问题。 (5)研究了分布在不同地理位置多制造商的生产运输协同调度问题,基于不同地理位置的多制造商等特点建立了相应的数学模型,目标函数为最小化整个制造过程的跨度时间。分析了该问题的最优解性质。提出了新颖的MGSA算法,设计了编码修正策略、初始化种群方法和最优种群保存策略,提出了插入操作、互换操作和变异操作。针对工件组批问题,构建了动态规划与启发式规则结合的DP-H混合算法。大量的仿真实验结果表明MGSA算法优于现有的GA和PSO算法。
[Abstract]:With the continuous development of network technology and global economic integration, the competition between supply chains is becoming more and more intense. More members of the supply chain are aware of the need to improve the competitiveness of the supply chain by strengthening the cooperation with other members, thus reducing the cost of their respective production and operation. The development of the Internet of things is just a supply chain. The cooperation between members provides the information base. It can not only feed the production and transportation information back to the management center of their own members in real time, but also share the information to other cooperative members in time. The technology of the Internet of things will push the cooperation among the members of the supply chain to a new level, and the use of these information can be used as a low production. Cost, increase profit and improve customer satisfaction, thus enhance the competitiveness of the overall supply chain, and further broaden the theoretical research field of production and transportation cooperation scheduling problem. Therefore, how to convert the value of the information into economic and social benefits, and to use the information and interest of the Internet of things to obtain efficient production and transportation cooperation scheme is the key to the problem. In this paper, based on the supply chain of aluminum products manufacturing, this paper studies the cooperative scheduling problem between the members of the supply chain in the environment of the Internet of things from the perspective of scheduling.
In this paper, the cooperative scheduling problem of multistage production and transportation in a variety of cases is systematically analyzed in the process of continuous batch processing in the extrusion plant. The finite situation of vehicles, the dynamic arrival of the workpiece, the deterioration of the working time of the workpiece, the failure of the machine and the multi manufacturer distributed in different geographical locations are considered respectively. Because these problems are all NP difficult, this paper focuses on the analysis of the properties of optimal scheduling schemes, and designs efficient heuristic and intelligent algorithms based on these properties. On the other hand, this paper derives the lower bounds for these problems. These lower bounds can be used to evaluate the accuracy of the algorithm. The research results of production and Transportation Co scheduling are summarized as follows:
(1) the cooperative scheduling problem of production and transportation under the limited condition of carrying vehicles is studied. The scheduling goal is to minimize the span of manufacturing span. A mathematical model is set up based on the limited conditions such as the limited transport vehicle. The problem is based on the relationship between the time of the batch and the processing time between the manufacturer and the manufacturer on the manufacturer's machine. It is divided into two cases. For two cases, two heuristic algorithms and the lower bounds are designed respectively. Based on the lower bounds of the problem, the worst case performance ratio of the heuristic algorithm is derived. A large number of small scale and large-scale random data experiments are designed considering the size of the workpiece. The simulation experiment results show that when the number of workpieces is 1000 The average relative difference between the two heuristic algorithms is 0.48% and 0.80% respectively, which are significantly better than the existing FOE and SPT algorithm and LOE and LPT algorithm.
(2) the cooperative scheduling problem of production and transportation is studied under the dynamic arrival of the workpiece. The scheduling goal is to minimize the span of manufacturing span. A mathematical model is established based on the constraint conditions of the dynamic arrival of the workpiece. The related properties of the dynamic arrival time of the workpiece in the optimal scheduling scheme are analyzed, and the optimal situation is discussed at the same time. The relationship between scheduling and batch number. Based on the properties of the above problems, a two stage TP-H heuristic algorithm is constructed, and the worst case performance ratio of the TP-H algorithm is 7/2., respectively, by relaxing the time of arrival of the work piece and assuming no extra idle time on the manufacturer's machine. Based on the different machine capabilities A large number of simulation experiments show the effectiveness of the TP-H algorithm. When the number of workpieces is 1000, the average relative gap of the proposed TP-H algorithm converges to 0.21%, which is superior to the existing MBF algorithm and the MEF algorithm.
(3) the problem of cooperative scheduling of production and transportation is studied. The problem of single machine scheduling in the case of workpiece deterioration is analyzed. The objective functions are to minimize the span time of manufacturing, minimize the number of delayed artifacts and minimize the total completion time of the chemical parts. The problem of cooperative scheduling of production and transportation in the case of cached area is considered, and a corresponding mathematical model is established. The objective function is to minimize the span time of manufacturing. The properties of batch and batch ordering in the optimal scheduling scheme are analyzed, and an optimization algorithm is constructed. In addition, the cooperative scheduling of production and transportation in the case of no cache area is also studied. A mathematical model based on no cache zone constraints is established. The objective function is to minimize the span of manufacturing span. The properties of the optimal solution are analyzed. A new lower boundary is derived and a heuristic algorithm is derived. The simulation experiment results show that the average and maximum relative gap ratio of the heuristic algorithm is less than 0.01 when the number of workpieces is greater than 260. It is.
(4) the problem of cooperative scheduling in production and transportation considering machine failure is studied. A mathematical model is established based on the possible failure of two parallel manufacturer machines. The objective function is to minimize the span time of manufacturing. The properties of the optimal scheduling scheme are analyzed and the lower bounds of the problem are derived. The scheduling rules under two circumstances are constructed, and the scheduling rules are constructed, respectively. These scheduling rules design a new heuristic algorithm. A large number of simulation results show that the heuristic algorithm can efficiently solve different scale problems.
(5) the problem of cooperative scheduling of production and transportation in different geographic locations is studied. A corresponding mathematical model is established based on the characteristics of multiple manufacturers in different geographical locations. The objective function is to minimize the span time of the whole manufacturing process. The optimal solution properties of the problem are analyzed. A novel MGSA algorithm is proposed and the design is designed. The coding correction strategy initializes the population method and the best species group preservation strategy, and puts forward the insertion operation, interchangeability operation and mutation operation. In view of the problem of the work group batch, a hybrid DP-H algorithm combining dynamic programming with heuristic rules is constructed. A large number of simulation results show that the MGSA algorithm is superior to the existing GA and PSO algorithms.

【学位授予单位】：合肥工业大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TP18;F273

【参考文献】