The methods of task pre-allocation and reallocation for multi-UAV cooperative reconnaissance mission

Nowadays, multi unmanned aerial vehicle (multi-UAV) systems have been widely used in battlefield. The rationality of mission plan can directly affect the effectiveness of multi-UAV system. The existing multi-UAV task allocation model lack a comprehensive modelling of task pre-allocation and task reallocation issues. However, in actual task execution, task pre-allocation and task reallocation are a holistic problem. Therefore, based on the background of multi-UAV cooperative reconnaissance, the authors establish a multi-UAV cooperative reconnaissance task pre-allocation and reallocation model (MCRTPR). There are two kinds of task allocation in MCRTPR model. One is task pre-allocation, which is a static task allocation before the mission begin. Another is task reallocation, that is a dynamic task allocation during the mission. For task pre-allocation, a particle swarm optimisation algorithm based on experience pool (EPPSO) is proposed. And for task reallocation, the authors design a partial task reallocation algorithm based on contract network protocol (CNP-PTR). The experimental results show that, compared with some state-of-the-art algorithms, EPPSO can get the lowest fitness value under various experimental conditions, and CNP-PTR is able to handle task reallocation problem caused by multiple kinds of dynamic events.