Social, economic and green optimization of the distribution process of e-commerce platforms
研究开发了一种多目标模拟退火算法，用于优化电商配送过程，同时考虑社会（司机代谢能耗）、经济（成本）和绿色（碳排放）可持续性，并比较了柴油车与电动车在不同场景下的表现。
During the last ten years, online shopping has continuously increased while embedding growing sustainability concerns regarding environment and, especially, drivers working conditions. Therefore, this paper presents a multi-objective simulated annealing (MOSA) developed to deal with a goods distribution problem characterized by social, economic and green sustainability aspects. This contribution compares three scenarios. The first one is distinguished by diesel vehicles and it neglects the load inside them. The second scenario considers the variation of the vehicle load along its route. Finally, the third scenario employs electric vehicles instead of diesel ones. The developed MOSA is implemented in real-world instances and results show that the load-based scenario performs similar to the one which ignores it, but it is more realistic since just 30% of the route is traveled with no load inside. In addition, the load-based scenario is more reliable since the metabolic energy consumption of the drivers depends also on this feature. Regarding this social aspect, the proposed contribution shows that the solution of the Pareto frontier which optimizes this aspect provides routes more balanced among drivers in terms of metabolic energy consumption, considering the personal characteristic of each operator. Furthermore, this paper indicates that the electric vehicles are more efficient, economically and environmentally, than diesel ones just in small areas. • Novel type of Vehicle Routing Problem with few pickup and many delivery nodes. • Social sustainability measured as metabolic energy consumption rate of drivers. • Multi-objective metaheuristic algorithm to solve a real-world distribution problem. • Carbon emissions of transports computed considering the dynamic load on the vehicle. • Comparison of diesel and electric vehicles on economic and environmental aspects.