An Ant Colony System Approach with Local Search for Multiobjective Logistics Scheduling

초록

Logistics scheduling arranges the delivery routes of employees, which can improve delivery efficiency. However, workload fairness is not sufficiently considered in existing re- search. In this article, a multiobjective logistics scheduling model is formulated, which simultaneously optimizes the total length of delivery routes and the fairness of workload distribution among delivery employees. In addition, real-world traffic data is incorporated into the proposed model to enhance its practicality. To deal with the proposed model, an ant colony system approach with local search (ACS-LS) is proposed. The ants in ACS- LS schedule the delivery routes of each employee and mainly optimize the total length of delivery routes. An archive is used to store the nondominated solutions found during the search process. To balance the working time among employees and optimize workload fairness, a local search strategy is designed in ACS-LS to adjust the assigned customers among delivery employees. 28 instances are constructed based on real-world traffic data in Guangzhou, China and the experimental results validate the effectiveness of the proposed ACS-LS approach.

키워드