Multi-UAVs delivery task decision based on Quint-Domain Interactive Multi-Objective Evolutionary Algorithm

This paper proposes a delivery task allocation model based on m-UAVs, aimed at maximizing the number of delivery tasks and minimizing the average and longest delivery times underling a fixed number of UAVs. A new discrete non- dominated sorting algorithm - the Quint-Domain Interactive Multi-Objective Evolutionary Algorithm (QDI-NSGA-II) is introduced to address the delivery task allocation problem. This algorithm is designed with a unique interactive mechanism of multi-population knowledge sharing based on the characteristics of the delivery task allocation model and improves upon the original mutation operations of GA-MTSP. Through comparison with NSGA-II and MOEA/D algorithms under different delivery task settings, the results show that QDI-NSGA-II offers a superior Pareto solution set for solving this problem.