The Formation Method of Task Coalition for Multi-UAV Cooperation

Due to the complexity of the environment and the weakness of capability of single UAV, the task coalitions are introduced into the multi-UAV cooperation. A task coalition model is established with the objective of the shortest mission time and the minimum coalition scale under the constraint of resources. An improved NSGA-II algorithm named VC-NSGA-II is proposed to form the task coalition by adding validity check and correction rules. It is effective to solve the problem of multi-resource tasks assignment by multi-heterogeneous UAVs. An improved K-means clustering algorithm is introduced into algorithm named Cluster-VC-NSGA-II under the situation of the large-scale tasks, while the tasks are divided into several task groups before task coalition. The efficiency of the VC-NSGA-II is compared with the NSGA-II algorithm by simulation. We also compared the VC-NSGA-II with the PGPA in different task assignment situations. The simulation results also demonstrated the clustering is useful to improve the efficiency of task coalition.