Heterogeneous UAV swarm grouping deployment for complex multiple tasks
A method of clustering before matching based on the improved K-means and deferred-acceptance(DA)algorithm is presented to solve the group deployment problem of heterogeneous unmanned aerial vehicle(UAV)swarm for complex multiple tasks.During the task clustering grouping stage,the approach of outlier detection and fixed initial cluster centers is exploited to increase the K-means clustering accuracy,and the grouping equalization adjustment strategy under margin is designed to enhance the grouping equalization based on the optimality condition.In the swarm grouping stage of matching,DA algorithm is developed by the preference list of task preferences to quickly generate a pre-selected scheme,and a two-stage conflict resolution is designed to ensure the matching stability and convergence.The simulation results show that the proposed method can solve the UAV swarm grouping deployment problem for complex multiple tasks quickly and effectively,and possess good optimality and timeliness.
万方数据
维普
