An Improved T-Distribution-Based Chaotic Black-Winged Kite Algorithm for Three-Dimensional Cooperative Path Planning of Multiple UAVs

Abstract

: To address the challenges of high search dimensionality, complex constraints, and susceptibility to local optima in multi-Unmanned Aerial Vehicle (UAV) cooperative path planning within complex 3D environments, an improved t-distribution-based chaotic Black-winged Kite Algorithm (ITCBKA) is proposed and applied to three-dimensional cooperative path planning for multiple UAVs. First, a multi-UAV three-dimensional cooperative path planning model is developed, and the fitness function is formulated by considering environmental constraints, trajectory quality, cooperative safety, and other relevant factors. Second, the Cubic chaotic map and opposition-based learning strategy are adopted to improve the distribution of the initial population. Meanwhile, an adaptive weight factor is incorporated to balance global exploration and local exploitation. In addition, a phase-adaptive t-distribution perturbation strategy and a differential perturbation strategy are integrated to strengthen global search ability and improve search accuracy in the later optimization stage. To validate the effectiveness of ITCBKA, its algorithmic performance is first evaluated on 20 benchmark test functions. Ablation experiments and statistical analyses further verify the effectiveness of the improvement mechanisms and the statistical significance of the performance differences. Then, ITCBKA is applied to cooperative path planning tasks involving two and three UAVs in two types of complex 3D scenarios. The results demonstrate that ITCBKA achieves superior optimization performance and stability and is capable of generating shorter, smoother, and more time-efficient flight trajectories under multiple constraints.