多目标低碳闭环供应链网络优化模型及算法

Multi-objective low-carbon loop supply chain network optimization model and algorithm

二氧化碳等温室气体排放的急剧增加,带来了日益严重的环境问题,低碳环保越来越受到重视。建立了一个多目标的低碳闭环供应链网络优化模型,模型以碳排放最小和成本最低作为优化目标。针对模型的特点,将遗传算法与ε约束法相结合编程,借助Sheffield大学的遗传算法工具箱和MATLAB 7.6平台,求得了双重目标条件下的Pareto非劣最优解,并验证了算法的有效性。与以往模型不同,该模型用维修中心、拆卸中心和分解中心代替再加工厂,细化了逆向供应链网络的主体。还探讨了废旧产品利用系数、分配系数对Pareto非劣最优解范围的影响,这种影响随系数的不同而不同,并分析了产生这种影响的原因。最后,研究了产品需求和废旧产品回收量需求随机时最优结果的变化范围,研究表明最优结果的范围变化不大。

In recent years,carbon dioxide and other greenhouse gas emissions has increased dramatically,which brings about seriously environmental problems.Low-carbon environment obtains more and more attention.This paper established a multi-objective low-carbon loop supply chain network optimization model.The optimization objectives of the model were to minimum carbon emissions and cost.According to the characteristics for the model,it combined genetic algorithm and ε constraint to program based on genetic algorithm toolbox of sheffield university and MATLAB 7.6 platform.It obtained the dual objective Pareto optimal solution.And it verified the effectiveness of the algorithm.Different from the previous models,this model substitutes reprocessing plants with repairing centers,dismantling centers and decomposition centers,which refines the subjects of reverse supply chain network.It discussed the impact of waste products utilization coefficient and distribution coefficient on the range of Pareto optimal solution.This impact varies with different coefficients.It also analyzed the reasons of this impact.Finally,it also studied the impact on the scope of optimal results in the conditions of random demand of products and random demand of waste products.Study shows little change in the range of optimal results.