《基于多目标遗传算法的预拌混凝土车辆调度》由会员分享，可在线阅读，更多相关《基于多目标遗传算法的预拌混凝土车辆调度（77页珍藏版）》请在金锄头文库上搜索。
1、上海交通大学硕士学位论文 1基于多目标遗传算法和仿真的预拌混凝土车辆调度 摘 要 水泥混凝土是当今世界用量最大、用途最广的建筑材料之一，年用量约 90 亿吨，是人类消耗除水以外最大的地球上的资源。国内外建筑工程所需的混凝土材料有很大一部份来自预拌混凝土供应厂， 预拌混凝土的发展是建筑工业化的重要标致。因此，全面使用预拌混凝土是我国乃至世界的趋势。 对于一个预拌混凝土厂来说，尽管原材料成本在总成本中占了很大的比例，但是控制预拌混凝土的配送成本对其利润来说更为重要。另外由于预拌混凝土本身的特点， 尤其是它的初凝对时间的限制非常严格，造成其配送调度是一个非常棘手的问题。目前，在预拌混凝土配送调度方面
2、，整个业界缺乏一套科学的调度方式，一般都是由调度员根据经验调度。因此，常常造成一方面部分有急需预拌混凝土的施工工地在急切等待混凝土预拌车配送预拌混凝土，另一方面，部分施工工地外混凝土预拌车排队等待卸料。 这种现象不但影响工地施工质量，同时也降低预拌混凝土厂的生产力和资源利用率，造成双方的损失。 本文结合利用了多目标遗传算法和计算机仿真技术初步解决了预拌混凝土配送调度问题。具体来说，首先笔者分析了预拌混凝土配送的特点和影响配送的因素等， 然后以此为基础为预拌混凝土的配送建上海交通大学硕士学位论文 2立了一个双目标调度模型， 该模型同时使施工工地空闲待料时间和混凝土预拌车在施工工地排队等待卸料时间
3、最短。 然后结合使用改进了的多目标遗传算法和仿真对调度模型进行求解。 在使用多目标遗传算法时，结合使用了自适应遗传算法和小生境技术对算法进行改进，另外，还对非支配解的构建策略、新群体的构建策略进行了比较好的设计，以此大大提高了多目标遗传算法的性能和效率。 以模型为基础，本文利用 VC工具开发出来了一个预拌混凝土配送调度系统（RMC-DSS） 。后面就详细介绍了如何利用该系统为预拌混凝土配送制定静态调度计划， 并且考虑到在实际的调度过程中可能出现的意外情况，本文还以系统为基础设计了预拌混凝土的“交互式调度”动态调度方案，当在实际配送过程中出现了不同的意外情况是，及时启用相应的动态调度方案，以避免
4、意外情况带来的损失。 总之， 利用 RMC-DSS， 一方面可以减少混凝土预拌车的等待时间，提高预拌混凝土供应厂的资源利用率；另一方面，可以减少施工工地的空闲等待时间，提高工地的工程质量。也就是说可以让预拌混凝土供应厂和施工工地达到双赢。 这对整个预拌混凝土行业甚至是建筑业来说都具有非常重大的意义。 关键字：预拌混凝土，多目标遗传算法，调度模型，帕累拖边界，静态调度，动态调度 知识水坝damdocdamdoc为您倾心整理(小店)(QQ2218108823)上海交通大学硕士学位论文 3OPTIMIZE THE SCHEDULE OF DISPATCHING RMC TRUCKS BASED ON
5、 MULTI-OBJECTIVE GA AND SIMULATION ABSTRACT Nowadays, the concrete is one of the most and widely used construction material, and is consumed about 9 billion tons per year, and which is the second widely used resource only inferior to water in the earth. The concrete used by the construction project
6、is usually supplied by the Ready-Mixed Concrete factory. The development of Ready-Mixed Concrete is a symbol of industrialization of construction. So comprehensively using the Ready-mixed concrete is the tendency of China and the whole world. For a Ready-Mixed Concrete factory, although raw material
7、s, such as sand, cement, and aggregates, present important costs to factory, the costs related to operating RMC trucks could be more critical to its revenues. Beside that, because of the special characteristic of Ready-Mixed Concrete, especially its strict time constraints, the dispatching of Ready-
8、Mixed Concrete truck is a very difficulty problem. Right now, the whole industry commonly distributes the Ready-Mixed concrete by the 知识水坝damdocdamdoc为您倾心整理(小店)(QQ2218108823)上海交通大学硕士学位论文 4experience of the dispatcher because it lacks an excellent scheduling model. So, we often found that when some s
9、ites were waiting for the Ready-Mixed concrete truck, many Ready-Mixed concrete trucks were queuing for unloading in other sites. It not only influences the quality of the construction project and lower the contractors profit, but also reduced the productivity of the Ready-Mixed concrete factory, wh
10、ich makes a lose-lose situation. This paper integrates Multi-Objective Genetic Algorithm and simulation technology to solve the problem of dispatching the Ready-Mixed Concrete trucks. In details, firstly, the characteristic of Ready-Mixed Concrete distribution and the factors that influence the dist
11、ribution were analyzed. Then, based on those, a double-objective dispatching model for Ready-Mixed Concrete distribution was set up, which minimizes both the trucks waiting time on the construction sites and the construction sites waiting time for Ready-Mixed Concrete trucks. And then, the dispatchi
12、ng model was solved by Multi-Objective Genetic Algorithm integrated with simulation technology. And the Multi-Objective Genetic Algorithm was optimized by integrated with adaptive Genetic Algorithm and niche technology. Whats more, the strategies that construct the non-dominated solutions set and ne
13、w generations were carefully designed, which can also improved the performance and efficiency of algorithm. 上海交通大学硕士学位论文 5Based on the dispatching model, a dispatching system for Ready-Mixed Concrete distribution, named RMC-DSS, was developed using VC+. Then the method of using the developed system
14、to schedule the dispatch plan for Ready-Mixed Concrete distribution was introduced. Whats more, considering the uncertainties or the accidents that might happen during the real distribution, the dynamic dispatch strategy was also designed based on the developed system. When a certain accident occurr
15、ed, a correspondent dynamic dispatch strategy was used in time, which can avoid the loss caused by the accident. In conclusion, for one hand, the dispatching system for Ready-Mixed Concrete can reduce the waiting time of Ready-Mixed Concrete trucks, which can improve the trucks utilization ratio; fo
16、r another, it can reduce the waiting time of construction sites, which can improve the project quality of construction. In one word, the research in this paper can get a win-win solution both for the Ready-Mixed Concrete factory and the construction sites. So this research has a high value to the Ready-Mixed Concrete industry and even the construction industry. KEY WORDS: Ready-mixed Concrete, Multi-Objective Genetic Algorithm, Pareto Font, Static D
