Self-adaptive metaheuristic-based emissions reduction in a collaborative vehicle routing problem

Abstract

Global climate change-related initiatives such as the 2015 Paris Agreement have highlighted the necessity of sustainable transportation. Nevertheless, the rapid growth of e-commerce has notably escalated vehicle kilometres travelled (VKT) and CO2 emissions within cities, posing a direct challenge to sustainability initiatives. To address these challenges, this study formulates a collaborative multi-depot green vehicle routing problem. This model utilises micro-consolidation centres (MCCs) as shared hubs alongside a microscopic approach linking emission rates to vehicle and route characteristics, in order to assess MCCs' effectiveness in reducing CO2 emissions. Introduced here is an innovative self-adaptive metaheuristic algorithm hybridising intelligent water drops and simulated annealing. This methodology differs from established approaches by incorporating a feedback control system that actively monitors the algorithm's performance and convergence towards the global minimum solution. Through continuous adjustments to algorithm parameters via a feedback loop, this methodology strikes a balance between exploitation and exploration. The algorithm is tested in a context-specific approach, first applying it to the Cordeau benchmark and comparing it with previous state-of-the-arts, followed by a case study comparing the collaborative network to an independent one. This approach achieves 43% and 25% reductions in VKT and emissions, respectively, enhancing urban logistics networks' efficiency and sustainability.