Shibutani, Rikuto (Nippon Institute of Technology), Kimura, Takayuki (Tokyo City University), Shindo, Takuya (Nippon Institute of Technology), Itoh, Nobuhiko (Nippon Institute of Technology)

Enhancing Grey Wolf Optimizer for the Traveling Salesman Problem Via Adaptive Neighborhood Learning and Dynamic Parameter Tuning

Scheduled for presentation during the Regular Session "S35a-Optimization, Control, and Learning for Efficient and Resilient ITS" (FR-LM-T35), Friday, November 21, 2025, 12:10−12:30, Surfers Paradise 2

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 18, 2025

Abstract

Efficient routing and delivery planning are crucial challenges in modern intelligent transportation systems (ITS), particularly in the context of smart logistics and urban mobility. In previous work, we proposed an Adaptive Large Neighborhood Search-Grey Wolf Optimizer (ALNS-GWO) as a heuristic for solving the Traveling Salesman Problem (TSP), a core combinatorial problem in transportation planning. However, the existing approach assigns uniform neighborhood operation weights and fixed control parameters across all agents, limiting solution diversity and adaptability. In this study, we propose an improved method, Adaptive Grey Wolf Optimizer (A-GWO), that introduces autonomous decentralized learning by assigning individual neighborhood preferences and dynamically tuning control parameters for each agent. Numerical experiments on TSPLIB benchmarks demonstrate that our method significantly outperforms the conventional approach in both accuracy and efficiency, especially for large-scale routing problems. The proposed A-GWO holds great promise for real-time vehicle routing and last-mile delivery optimization in ITS applications.