Electric vehicle routing optimization for sustainable kitchen waste reverse logistics network using robust mixed-integer programming

This paper proposes an innovative reverse logistics network (RLN) to manage kitchen waste (KW) transportation and resource treatment. The network employs battery electric (BE) trucks for transportation, and the challenge lies in determining the distribution of various KW treatment centers and establishing the optimal transportation routes for KW and its residues. The proposed RLN is self-sufficient, because the electricity produced by the centers within the network is adequate to power the BE trucks. We develop a matched mixed-integer programming model to optimize the entire process, with the goal of minimizing the total potential economic and environmental costs. Notably, the model considers comprehensive cost components and employs a carbon trading policy to translate carbon emissions into carbon costs. We use robust optimization to generate optimal solutions that remain viable even under the worst-case scenario concerning uncertain parameters. We then test the feasibility of the proposed methodology in a real-world case. We conduct specific scenario analyses on capacity and mode of trucks to offer practical KW transportation strategies and recommendations. We found that the larger the capacity of a BE truck, the greater the economic and environmental benefits for the KW RLN. The self-sufficient KW RLN using BE trucks proved to be the least costly, followed by the ordinary KW RLN using BE trucks, while the KW RLN using diesel trucks was the most expensive and environmentally detrimental.