Abstract:

Due to the diversity of platform characteristics, intelligent unmanned clusters have good resource allocation space and functional flexibility, and can respond to complex and ever-changing task requirements. However, most existing researches have not considered practical needs such as the heterogeneity and correlation of tasks, and their task allocation methods still lack adaptability to the requirements and resources, as well as dynamic response capabilities for cluster collaboration. This paper proposed a dynamic response overlapping coalition task allocation architecture for unmanned cluster systems to address problems such as uneven task allocation, poor collaboration, and weak dynamic adaptability in the process of performing multi-objective tasks. The architecture considered the value, priority, and requirements of multiple coupled heterogeneous tasks, as well as the impact of task changes, to construct a coalition game model and designed a coalition formation algorithm. It distributed and coordinated the resources of different cluster members, including unmanned aerial vehicles and unmanned ground vehicles, to achieve a reasonable match between heterogeneous member resources and heterogeneous tasks. Furthermore, it can efficiently perform dynamic adjustments according to changes in tasks. Simulation results show that the proposed algorithm can adapt to dynamic task scenarios, form stable and efficient task coalitions and resource allocation results, improve the system benefits and success rate of unmanned clusters performing diverse heterogeneous tasks, and achieve collaborative task allocation optimization under dynamic conditions.

Key words: unmanned clusters, intelligent decision-making, overlapping coalition formation game, dynamic task assignment