考虑选址-路径-库存联合优化的碳排放多目标模型与算法

Considering the Location-routing-inventory Joint Optimization of Carbon Emissions Multi Objective Model and Algorithm

研究了存在拥堵和限速的路况依赖下,同时考虑选址-路径-库存联合优化和碳排放的多目标模型。根据道路类型和行驶状态的不同,将限速和拥堵因素嵌入到模型的构建中。对限速函数的的设置以社会福利最大化为基准,对拥堵问题给出了4级拥堵设置,通过0-1变量对其进行规范化处理。通过多目标优化模型来权衡总成本和碳排放量的影响,使用标准化正规化方法(NNC)求解和优化帕累托解集,基于中石油东北化工销售公司的计算实验与分析表明,所构建的模型是有效的。

There is a growing concern over carbon emissions issues in supply chain management. Compulsory green policies have resulted in changes to supply chain management for companies participating in the carbon abatement or reducing carbon emissions. This paper examines the relationship between economic cost and carbon emissions in collaboration location-routing-inventory problems （CLRIP） under the road dependence condition. Previous research has not examined legislations and other exogenous variables, such as carbon price. We make a Wade-off between economic cost and carbon emissions through multi-objective optimization. The Pareto fi＇ontier result can provide optimal solutions according to decision makers＇ preference. In addition, limiting vehicle speed and congestion are incorporated into our models. Firstly, some assumptions about limiting vehicle speed and congestion arc proposed. One assumption is that population distribution is incorporated into the function of limiting vehicle speed, which takes into account both equity and efficiency. The other assumption is that we propose four levels to identify the congestion situation, and the 0-1 binary code is used to normalize the variable. In addition, a multi-objective mixed-integer formulation for the CLRIP is provided. To our best knowledge, this model is the first that incorporates carbon emissions into the CLRIP without any exogenous variables. The multi-objective mixed integer programming model explicitly considers the environmental issue by introducing the limiting vehicle speed and congestion. Thirdly, a normalized normal constraint （NNC） method is applied to solve the multi-objective problem. The NNC method is a posteriori articulation ofpreference method and can find a set of evenly distributed Pareto solutions so that the results can provide decision makers configuration portfolios. Lastly, we conduct a case study based on the source of China National Petroleum Corporation （CNPC）, a northeastern chemical sales company. The numerical results characterize the frontier of the non-dominated set and also show sensitivy analysis of some critical parameters in the model. Consequently, some useful managerial insights are learned. For instance, it is possible to significantly reduce carbon emissions without significantly increasing cost by adjusting the ratios of speed limit and congestion at a certain range. To summarize, this paper uses a multi-objective optimization method to study the CLRIP with carbon emissions considering road dependence. Major findings of this paper can provide some valuable theoretical and practical implications for decision makers to make effective decisions in supply chain operations under strict carbon emissions policy.