Abstract
To address multi-UAV cooperative path planning in complex three-dimensional dynamic environments, this paper proposes an Improved Spider Wasp Optimization (ISWO) algorithm to overcome the original SWO’s shortcomings: poor population diversity, low convergence accuracy, premature convergence, and weak dynamic obstacle avoidance. Three key optimizations are introduced: a Logistic–Tent hybrid chaotic map initializes the population to expand the search range and improve diversity; modified chasing and escaping update strategies enhance adaptive search and global exploration; an optimal neighborhood perturbation mechanism balances exploration and exploitation, improving local optima escape and convergence stability. A multi-constraint 3D path planning cost model is constructed, considering path length, smoothness, altitude, energy consumption, and threat avoidance to meet practical UAV flight requirements. Experiments on static, dynamic threat, and multi-UAV cooperative scenarios (5 and 7 UAVs) compare ISWO with the original SWO and mainstream algorithms, with ablation experiments verifying each improvement. Results show that ISWO achieves better accuracy and stability in static scenarios, reduces path cost and collision rates in dynamic environments, and generates shorter, smoother, safer trajectories, demonstrating its excellent robustness and adaptive planning capability in complex 3D environments.