Multi-Type UAVs Cooperative Task Allocation Under Resource Constraints

Abstract

Coordinated task allocation for multiple unmanned aerial vehicles (multi-UAVs) is an important problem. Taking considerations of the types of UAVs, and the resources are extremely significant in the coordinated control of multi-UAVs. In the interests of assigning tasks efficiently and accurately for the cooperative UAVs of different types, the advanced multi-UAVs control technology requires a universal task assignment method under resource constraints. In this paper, we introduce a novel multi-type UAVs coordinated task allocation method based on cross-entropy (CE), and take the resources required for tasks into account. The CE method takes random samples from the candidate solutions, and then uses them to update the allocation probability matrix. We address the specific processes of CE dealing with the constrained multi-type UAVs task allocation problem, and reveal that CE has the advantage of solving large scale allocation problems. Furthermore, numerical simulations of CE handling task assignment, and comparisons with the exhaust search method are conducted to validate the merits of the cross-entropy method dealing with the considered problem.