A Dual-Encoding-based Genetic Algorithm for Multi-Objective Multi-UAV Scheduling in Firefighting Scenarios
Meng Gao, Xiao-Fang Liu, Zhi‐Hui Zhan, Jun Zhang · 2025
Unmanned aerial vehicles (UAVs) are increasingly used for firefighting due to security. Multiple UAVs depart from a site and cooperate to execute tasks with time-varying demands in different locations. To better execute firefighting tasks, the reasonable scheduling of UAVs is crucial. Although multiple methods have been developed to solve the multi-UAV scheduling problem, they mainly focus on instances with sufficient UAV resources. They face challenges on instances with limited resources and multiple objectives in terms of solution diversity and convergence, especially when the number of tasks is larger than that of UAVs. This paper models the problem as a bi-objective optimization problem, aiming to minimize two important objectives: the makespan and the total travel distance of UAVs. A dual-encoding-based genetic algorithm (DEGA) is developed to solve the problem. In DEGA, a dual-encoding scheme is adopted for solution representation, in which the execution order of all tasks is represented as a sequence and the task assignment for UAVs is represented as a binary matrix. Correspondingly, solutions can be constructed in two steps: task sequence generation first and then task assignment. Crossover and mutation operations are specifically designed to explore the search space for enhancing solution diversity. In addition, a local search is employed to improve solution quality. Experimental results on instances with various scales demonstrate that DEGA outperforms state-of-the-art algorithms on most instances in terms of solution optimality and diversity.