Multi-UAV Cooperative Search for Moving Targets With Impaired Communication Using Improved Gray Wolf Optimizer

Abstract

The study investigates the problem of collaborative target search by multiple unmanned aerial vehicles (UAVs) under the condition of communication information corruption. The problem involves the influence of target movement and environmental obstacles on the update of the UAV’s local target probability map (TPM), as well as communication information corruption among UAVs due to packet loss and resulting local information inconsistency. The objective is to efficiently plan paths for the UAV swarm to discover more targets. To address the above issues, this paper proposes the following methods. Firstly, an information diffusion model is introduced to simulate the TPM changes caused by target movement and the discovery of obstacles based on a two-dimensional Gaussian distribution, thereby correcting the real-time target information held by UAVs. Secondly, a radial basis function neural network (RBFNN) is employed to recover the corrupted information, and a weighted averaging method is used to fuse the information from multiple UAVs, achieving consensus among the UAVs. Lastly, the original gray wolf optimizer (GWO) is enhanced by incorporating strategies such as Levy flight and the nonlinear convergence factor. Subsequently, the improved GWO is integrated with the model predictive control (MPC) for UAV trajectory planning, enabling a more efficient search for optimal paths. Experimental results demonstrate that the proposed information recovery method outperforms the compared methods by 24.7% in terms of performance. The proposed comprehensive search approach shows higher efficiency in searching for more targets within a specified time, and it also exhibits advantages in terms of area coverage and other aspects.