Decentralized Task Assignment of Multiple UAVs Based on the Communication Strategy for Dynamic Environments

This paper deals with a task assignment problem of the multiple autonomous unmanned aerial vehicles (UAVs). The task assignment problem is one of the decision making procedures of the UAV group, and it has been interpreted as a combinatorial optimization problem. Task assignment problem has been usually solved in centralized manner for the optimization aspect, but it is not suitable for the dynamic environment and the autonomous process of UAVs because of the computational and communication loads. Therefore, decentralized or distributed task assignment process is required. In this study, a real-time decentralized task assignment method is proposed for autonomous UAVs. Unlike the centralized task assignment, it requires the consideration about communications among UAV groups particularly to deal with the dynamic environment. The proposed method is divided into two stages: the order optimization stage and the communications and negotiation stage. Genetic algorithm and negotiation strategy based on one-to-one communication is adopted for each stage. Through the proposed method, a non-conflicting and real-time decentralized task assignment of the multiple UAVs can be performed. The validity of the method is evaluated through the numerical simulations.