Distributed coordinated task allocation for heterogeneous UAVs based on capacities

Abstract

This paper investigates the problem of allocating unmanned aerial vehicles (UAVs) with different capacities to targets performing detection, attack and verification tasks. One of characteristics of capacities of UAVs or capacities required by targets is described by the concept of generalization, and modeled utilizing the normalized entropy. The discrepancy in capacities between two UAVs is represented by capacity distance, and measured by extending relative entropy. The model of the distributed task allocation for heterogeneous UAVs is established based on the principles that: (1) the capacities between UAV own and required by the target should be matched; (2) all the three tasks on one target should be executed by one UAV as possible; (3) the latter phase task of a target is tend to be executed by the UAV with large capacity distance from the UAV which has performed the former phase task of that target; (4) the route that the UAV traveled should be short and the flight range constraint should be satisfied. And the distributed auction algorithm using limited communications is developed to solve the task allocation problem. Finally, simulation results demonstrate the reasonableness of the model and the effectiveness of the method.