城市固体废弃物运输和处理设施选择的鲁棒优化:基于ROME方法  

Robust optimization model for transportation and processing facilities selection in municipal solid waste system: A rome method

作　　者：秦全德[1] 李丽[1] 程玉荣[1] QIN Quan-de;LI Li;CHENG Yu-rong(College of Management,Shenzhen University,Shenzhen 518060,China)

机构地区：[1]深圳大学管理学院

出　　处：《管理工程学报》2019年第3期179-187,共9页Journal of Industrial Engineering and Engineering Management

基　　金：国家自然科学基金资助项目(71402103)

摘　　要：针对城市固体废弃物产生量及处理设施成本不确定的特点,建立了城市固体废弃物管理系统的鲁棒优化模型。由于难于运用概率密度函数形式刻画城市固体废弃物管理系统中的不确定性,提出了非概率鲁棒方法处理该系统中的运输和处理设施选择问题。针对构建模型求解的复杂性,根据Joel和Melvyn(2011)设计的ROME(Robust Optimization Made Easy)方法,采用基于 ROME 的相关决策准则简化模型结构,从而对构建的鲁棒优化模型进行求解。数值例子的实验结果表明了提出的模型和方法在城市固体废弃物管理系统应用中的有效性。Due to rapid industrialization and increasing urban population in China, the amount of municipal solid waste (MSW) has increased dramatically. Currently, MSW has serious impact on public health and ecological environment in urban cities. The challenges China must face with regards to waste management have been a top concern to government, public and scholars. MSW management is complicated with a variety of uncertainties which is associated with waste generation, transportation, treatment and disposal processes. To plan an efficient MSW system, the uncertain mathematical programing methods have been widely adopted for dealing with uncertainties. Robust optimization is a relatively new stochastic method for uncertain optimization problems. Compared to other uncertain optimization technologies, robust optimization not only can deal with uncertainties, but also can determine the system stability. When uncertainties of parameters cannot be defined in the form of a probability density function, non-probabilistic robust optimization can be a good alternative. Non-probabilistic robust optimization model does not need to know the probability density distribution function to define the uncertain parameters or artificially set the scenario probability. Thus, it is more suitable to the practical situation. Robust optimization problems are difficult to solve. Generally, they should be converted into the robust counterpart, and then use standard optimization algorithms to solve. During the robust counterpart transforming process, it needs to add additional variables, increasing the complexity and changing the structure of the original model. Considering the uncertainty of waste amount generated and operating costs in MSW system, this paper developed a robust optimization model to handle the problem of transportation and processing facilities selection. As it is difficult to describe the uncertainty of waste amount and operating costs using probability theory in MSW management, a non-probabilistic robust optimization model is e

关 键 词：城市固体废弃物 不确定性 非概率鲁棒优化 ROME方法 

分 类 号：F224[经济管理—国民经济]