基于多目标优化的鲁棒双行设备布局方法研究
发布时间：2018-10-17 12:54
【摘要】：随着社会竞争日益激烈,越来越多的企业改粗放型生产方式为集约型生产方式,大量采用先进生产设备。设备布局作为企业规划的重要部分之一也越来越受到制造业的重视。本文以工厂生产车间设备布局为研究对象,分析设备布局研究现状、设备布局特点,研究以设备占地面积和多个生产阶段里设备之间总物流成本为优化目标的鲁棒双行设备布局方法。 目前学者研究多行和单行设备布局偏多,对于双行设备布局研究甚少,而实际工厂中双行布局也具有很大实际意义。考虑到问题的特殊性,本文着眼于双行设备布局问题。另外,很多前人学者的文献大多基于单生产阶段的布局问题研究。随着实际生产线的需要,物流量往往随着不同的生产阶段变化。当设备布局改变时,企业其实更关心在多个阶段里的最后布局总开销。静态布局往往不能适应市场需求,而动态布局又存在重置布局的成本问题,因此本文将多个生产阶段纳入研究且采用鲁棒设备布局。 针对鲁棒双行设备布局问题,本文提出一种基于分解的多目标进化算法(MOEA/D算法)的求解方法,主要做了如下几方面的工作：1)采用Pareto最优解方法,优化鲁棒双行设备布局问题中的多阶段总物流成本和设备占地面积两个优化目标。2)采用带拥挤度算法的MOEA/D算法用于求解鲁棒双行设备布局问题中的连续问题。3)采用切比雪夫分解法将多目标问题分解成一系列子问题进行优化。4)鲁棒双行设备布局问题同时包含离散问题和连续问题,本文研究了MOEA/D算法同时求解这两个问题以及单步求解这两个问题时的区别。5)小规模问题通过与CPLEX对比研究MOEA/D的有效性,大规模问题通过实验结果分析算法的稳定性和有效性。 通过研究本文得出MOEA/D算法在求解鲁棒双行设备布局问题时具有效性、准确性和稳定性。
[Abstract]:With the increasingly fierce social competition, more and more enterprises change the extensive mode of production to intensive mode of production, and adopt a large number of advanced production equipment. As an important part of enterprise planning, equipment layout is paid more and more attention by manufacturing industry. In this paper, the factory workshop equipment layout as the research object, analysis of the status quo of equipment layout, equipment layout characteristics, In this paper, a robust two-row equipment layout method is studied, in which the equipment covers an area and the total logistics cost between the equipments in multiple production stages is taken as the optimization objective. At present, many scholars study the layout of multi-row and single-row equipment, but there is little research on the layout of two-row equipment, and the dual-row layout in the actual factory is of great practical significance. Considering the particularity of the problem, this paper focuses on the layout of two-line equipment. In addition, many previous scholars mostly based on the single-stage layout of the study. With the demand of the actual production line, the material flow often changes with different production stages. When the device layout changes, the enterprise is actually more concerned with the final layout overhead in multiple phases. Static layout is often unable to meet the market demand, and dynamic layout has the cost of resetting layout. Therefore, in this paper, multiple production stages are included in the study and robust equipment layout is adopted. In this paper, a decomposition based multi-objective evolutionary algorithm (MOEA/D) is proposed to solve the robust two-row device layout problem. The main works are as follows: 1) the Pareto optimal solution method is adopted. There are two optimization objectives in the robust two-row equipment layout problem: the total logistics cost of multi-stage and the space occupied by the equipment. 2) the MOEA/D algorithm with congestion algorithm is used to solve the continuous problem in the robust two-row equipment layout problem. 3) the continuous problem in the robust two-row equipment layout problem is solved by using the MOEA/D algorithm with congestion algorithm. The method of Chebyshev decomposition is used to decompose the multi-objective problem into a series of sub-problems. 4) the robust two-row layout problem includes both discrete and continuous problems. In this paper, we study the difference of MOEA/D algorithm in solving these two problems simultaneously and in one step. 5) small scale problems are compared with CPLEX to study the effectiveness of MOEA/D, and large scale problems are analyzed by experimental results to analyze the stability and effectiveness of the algorithm. By studying the MOEA/D algorithm, it is concluded that the algorithm is effective, accurate and stable in solving the robust two-line device layout problem.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP18
本文编号：2276723
[Abstract]:With the increasingly fierce social competition, more and more enterprises change the extensive mode of production to intensive mode of production, and adopt a large number of advanced production equipment. As an important part of enterprise planning, equipment layout is paid more and more attention by manufacturing industry. In this paper, the factory workshop equipment layout as the research object, analysis of the status quo of equipment layout, equipment layout characteristics, In this paper, a robust two-row equipment layout method is studied, in which the equipment covers an area and the total logistics cost between the equipments in multiple production stages is taken as the optimization objective. At present, many scholars study the layout of multi-row and single-row equipment, but there is little research on the layout of two-row equipment, and the dual-row layout in the actual factory is of great practical significance. Considering the particularity of the problem, this paper focuses on the layout of two-line equipment. In addition, many previous scholars mostly based on the single-stage layout of the study. With the demand of the actual production line, the material flow often changes with different production stages. When the device layout changes, the enterprise is actually more concerned with the final layout overhead in multiple phases. Static layout is often unable to meet the market demand, and dynamic layout has the cost of resetting layout. Therefore, in this paper, multiple production stages are included in the study and robust equipment layout is adopted. In this paper, a decomposition based multi-objective evolutionary algorithm (MOEA/D) is proposed to solve the robust two-row device layout problem. The main works are as follows: 1) the Pareto optimal solution method is adopted. There are two optimization objectives in the robust two-row equipment layout problem: the total logistics cost of multi-stage and the space occupied by the equipment. 2) the MOEA/D algorithm with congestion algorithm is used to solve the continuous problem in the robust two-row equipment layout problem. 3) the continuous problem in the robust two-row equipment layout problem is solved by using the MOEA/D algorithm with congestion algorithm. The method of Chebyshev decomposition is used to decompose the multi-objective problem into a series of sub-problems. 4) the robust two-row layout problem includes both discrete and continuous problems. In this paper, we study the difference of MOEA/D algorithm in solving these two problems simultaneously and in one step. 5) small scale problems are compared with CPLEX to study the effectiveness of MOEA/D, and large scale problems are analyzed by experimental results to analyze the stability and effectiveness of the algorithm. By studying the MOEA/D algorithm, it is concluded that the algorithm is effective, accurate and stable in solving the robust two-line device layout problem.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP18
【参考文献】
相关期刊论文 前9条
1 李英娜,李建勇,李志辉;柔性制造系统鲁棒性设备布局研究[J];北方交通大学学报;2001年04期
2 陈皓;崔杜武;严太山;李凌波;;基于竞争指数的模拟退火排序选择算子[J];电子学报;2009年03期
3 黄卫华;许小勇;范建坤;;实数编码遗传算法中常用变异算子的Matlab实现及应用[J];广西轻工业;2007年01期
4 郝国生;严玉若;黄永青;胡培亮;章磊;路娟;季君;;基于三角函数的遗传算法选择算子[J];江南大学学报(自然科学版);2010年02期
5 文诗华;郑金华;李密青;;多目标进化算法中变异算子的比较与研究[J];计算机工程与应用;2009年02期
6 郑向伟;刘弘;;多目标进化算法研究进展[J];计算机科学;2007年07期
7 刘琼;许金辉;张超勇;;基于改进蛙跳算法的鲁棒性车间布局[J];计算机集成制造系统;2014年08期
8 公茂果;焦李成;杨咚咚;马文萍;;进化多目标优化算法研究[J];软件学报;2009年02期
9 李晨;宁红云;;改进的遗传算法选择算子[J];天津理工大学学报;2008年06期
，本文编号：2276723
本文链接：https://www.wllwen.com/guanlilunwen/wuliuguanlilunwen/2276723.html
