可持续逆向物流网络设计研究进展及趋势

Progress and prospects of sustainable reverse logistics network design

随着可持续发展理念的深入践行以及逆向物流活动对环境与社会影响的日益显著,可持续逆向物流网络设计问题正成为研究的热点.首先系统总结了可持续逆向物流网络研究的总体现状,并对经济、环境、社会目标测度指标与方法、数学模型的决策变量、目标函数与约束条件、不确定因素考虑及处理方法、问题求解方法与工具以及研究应用情况进行概述.研究发现,经济、环境与社会影响测度指标不完善,尤其是环境与社会影响类决策变量考虑较少,社会可持续研究总体偏少且测度方法有待突破.同时,较少考虑多周期、多产品、越库作业、按质量等级回收、不确定环境等真实逆向物流运作场景,战略、战术与运作层面集成优化不够,以上这些均限制了模型的应用.面对复杂的数学模型,较少采用区间优化、模拟仿真等方法求解不确定问题,更为高效、可靠的元启发式算法或精确算法亟待设计,同时应用仿真优化技术、交叉熵等进行效能评估或算法间比较以提升网络设计的效率和科学性.最后展望了可持续逆向物流网络设计潜在的发展趋势.

With the deepening practice of the concept of sustainable development and the increasingly significant impact of the environment and society of reverse logistics activities,the design of sustainable reverse logistics network is becoming a research hotspot.Firstly,this paper systematically summarizes the general situation of sustainable reverse logistics network design research.Then the paper overviews the progress from five aspects,including the indicators and methods for measuring economic,environmental and social sustainability,the decision variables,objective functions and constraints of mathematical models,the uncertainties considered and the treatment techniques,the methods and tools of model solving and research application.It is found that the indicators of economic,environmental and social impact measurement need to be expanded,especially the decision-making variables of environmental and social impact are less considered,the overall research of social sustainability is less and the measurement method needs to be broken through.At the same time,the real reverse logistics operation scenarios such as multi-period,multi product,cross docking,collecting according to quality level and uncertain environment,are not considered enough,and the integration and optimization of strategies,tactics and operation levels are not enough.All of above restrict the application of the model.In the face of complex mathematical models,interval optimization and simulation are seldom used to solve uncertain problems,theremore,more efficient and reliable meta heuristic algorithms or precise algorithms need to be designed urgently.Moreover,simulation optimization technology and cross entropy are used for efficiency evaluation or algorithm to algorithm ratio to improve the efficiency and scientificity of network design.Finally,the prospects of sustainable reverse logistics network design are presented.