【作者】 李蓉蓉；

【导师】 胡天军；

【作者基本信息】 北京交通大学 ， 交通运输规划与管理， 2008， 硕士

【摘要】 逆向物流是伴随着可持续发展和客户满意度提高而产生和发展的,随着环境保护重视程度的增大和废旧资源再利用价值的明显,许多企业都纷纷构建自己的逆向物流体系,回收废旧资源用以创造新的经济效益。逆向物流网络是整个逆向物流体系运作的基础和关键,经济、高效的逆向物流网络是许多企业构建的目标。因此,如何构建出合理的逆向物流网络在逆向物流领域具有实际的研究价值。本文正是基于以上背景,在国内外对逆向物流网络研究成果的基础上,从确定性和不确定性两个方面对逆向物流网络进行了建模和优化。文章从逆向物流的概念着手,综合了国内外对逆向物流的定义,系统地诠释了逆向物流的内涵。之后,对逆向物流网络进行了系统的分析,从八个不同的角度对逆向物流网络的类型进行了划分,为网络的构建提供了基础和前提。本文的核心部分在于逆向物流网络模型的构建,文章构建了两个模型。第一个模型研究了确定情况下,由企业主导的多产品、多层次的正、逆整合型网络模型,该模型融合了开环和闭环的两种回流模式,模型设置了回收点、再处理中心、资源再生中心和再制造厂四级处理设施,同时考虑到再制造和再循环两种类型的回流特点,并设计了相应的遗传算法求解;第二个模型研究了数量不确定情况下的单产品、开环型逆向物流网络模型,模型设置了回收点和再处理中心两级处理设施,并且设计了由遗传算法、Monte Carlo模拟、线性规划相结合的混合智能算法进行求解。两个模型均给出算例对算法的有效性进行了验证,并对比了设计算法与传统优化方法的求解结果,同时对智能优化方法的参数进行了灵敏度分析。最后,对本文的研究内容和下一步的研究方向作出了总结和展望。更多还原

【Abstract】 Reverse logistics is developed with sustainable development and servings. With the increased environmental concerns and evident value of returns, plenty of enterprises started their reverse logistics career for gaining more profit. Reverse logistics network is the base and sticking point, economical and efficient network is the goal of most enterprises. Therefore, how to design an appropriate reverse logistics network has practical value on research.In the view of this, this paper design and optimize reverse logistics networks on certain and uncertain conditions. First, Chapter one and two put forward a new concept of reverse logistics, and plot out eight types of network. Second, Chapter there designs a network model with certain condition for enterprise with multi-product, multi-hierarchy and two flows of logistics, this model contains four hierarchies which is collection site, reprocessing centre, recycling centre, remanufacturing factory, and includes functions of remanufacturing and recycling, then optimize the model with Genetic Algorithms. Chapter four designs a network model with uncertain condition for open-loop network, contains two hierarchies of collection site and reprocessing centre, then optimize the model with mixed intelligent optimization composed by Genetic Algorithms, Monte Carlo simulation and linear programming. Both models give a case study, compare the result of traditional optimization and intelligent optimization, and analyze diversification on parameters. Finally, Chapter five sums up content and deficiency of paper.更多还原