Optimization of Charging-Station-Assisted Multi-UAV Data Collection Task Allocation

Abstract

In the field of mobile crowd sensing (MCS), unmanned aerial vehicles (UAVs) provide an efficient platform for large-scale and distributed data collection. However, their limited onboard energy restricts mission coverage in large-scale or remote areas. To address the task allocation problem in multi-UAV data collection scenarios and meet the multi-objective optimization requirements of task coverage, data throughput, and energy consumption within the target area, this paper develops an integrated mathematical model of UAV energy consumption and charging, and proposes an improved charging-aware NSGA-II algorithm framework. The proposed algorithm can insert charging stations into UAV routes to restore solution feasibility, while retaining solutions with higher task coverage through a multi-objective dominance rule. Experimental results show that the on-demand charging repair mechanism can effectively improve route feasibility and service coverage under energy-constrained scenarios. In several scenarios with comparable coverage rates or the same amount of collected data, NSGA-II-SF reduces operational energy consumption compared with the NN and KMNN baselines, and achieves a reasonable multi-objective trade-off among service coverage, data collection volume, and energy consumption.