2952. A Hybrid Metaheuristic and Column Generation Approach for Efficient Urban Electric Vehicle Routing
Invited abstract in session MC-58: Electric Vehicles, stream Vehicle Routing and Logistics.
Monday, 12:30-14:00Room: Liberty 1.13
Authors (first author is the speaker)
| 1. | YUTONG QI
|
| University of Manchester |
Abstract
With digital transformation and growing environmental awareness, electric vehicles have become a key part of urban mobility, offering great potential for reducing carbon emissions and boosting energy efficiency. However, battery limitations and charging constraints make efficient route planning a major challenge in large-scale operations. To tackle this, we propose a novel approach that combines a hybrid metaheuristic with column generation techniques to optimize electric vehicle routing. Our method uses an adaptive neighborhood adjustment operator to capture dynamic routing changes, replacing conventional fixed local search methods. We also employ column generation to decompose the problem into master and sub-problems, generating only the necessary candidate routes and using dual information for local refinement. Moreover, a dynamic candidate set is maintained by continuously adding new solutions and removing redundancies, which helps ensure diversity and stability. Experiments with real-world data from Chinese cities show that our approach not only improves solution accuracy and stability but also reduces computation time. This makes it an efficient and practical optimization pathway for urban electric vehicle dispatch.
Keywords
- Vehicle Routing
- Column Generation
- Metaheuristics
Status: accepted
Back to the list of papers