Key points are not available for this paper at this time.
In the task assignment problem of multi-UAV collaborative reconnaissance, existing algorithms have issues with inadequate solution accuracy, specifically manifested as large spatial spans and knots of routes in the task execution of UAVs. To address the above challenges, this paper presents a multi-UAV task assignment model under complex conditions (MTAMCC). To efficiently solve this model, this paper proposes an elite wolf pack algorithm based on probability threshold (EWPA-PT). The EWPA-PT algorithm combines the wandering behavior in the traditional wolf pack algorithm with the genetic algorithm. It introduces an ordered permutation problem to calculate the adaptive wandering times of the detective wolves in a specific direction. During the calling phase of the algorithm, the fierce wolves in the wolf pack randomly learn the task assignment results of the head wolf. The sieging behavior introduces the Metropolis criterion from the simulated annealing algorithm to replace the distance threshold in traditional wolf pack algorithms with a probability threshold, which dynamically changes during the iteration process. The wolf pack updating mechanism leverages the task assignment experience of the elite group to reconstruct individual wolves, thereby improving the individual reconstruction’s efficiency. Experiments demonstrate that the EWPA-PT algorithm significantly improves solution accuracy compared to typical methods in recent years.
Zhang et al. (Sun,) studied this question.