A Multi-Strategy Ant Colony System for Multi-Objective Multi-UAV Path Planning Considering Safety Risk and Delivery Lateness

초록

Multi-UAV delivery in urban environments faces critical challenges arising from the trade-off between delivery efficiency and operational safety under strict payload and battery constraints. To address this issue, this paper formulates CMURP model as a bi-objective optimization model that simultaneously minimizes total delivery lateness and cumulative safety risk. The model explicitly incorporates payload capacity limits, energy consumption, and environmental risk, revealing the inherent conflict between shortest-path efficiency and risk avoidance. To solve the problem, a Multi-Objective Multi-Strategy Ant Colony Optimization algorithm is proposed. The framework integrates a simulation-based evaluation process to strictly enforce feasibility, and three pheromone evolution strategies to respectively emphasize convergence, safety-efficiency balance, and Pareto front diversity. Extensive experiments under varying spatial scales, workload intensities, and proportional expansions demonstrate that the proposed method achieves strong convergence, high solution quality, and robust performance under resource-saturated scenarios. In particular, the diversity-preserving strategy maintains non-zero service quality at critical load levels and exhibits superior stability in complex environments. The results confirm that the proposed MSACO provides an effective and scalable solution framework for safety-aware multi-UAV logistics planning in urban delivery systems. © 2026 IEEE.

키워드