UAV path planning based on improved Whale Optimization Algorithm in multi-obstacle scenarios

Abstract

To address the issues of poor convergence performance and susceptibility to local optima in the traditional Whale Optimization Algorithm (WOA) for 3D path planning of unmanned aerial vehicles (UAVs), we propose an improved Whale Optimization Algorithm (R*WOA) that integrates the Rapidly Expanding Random Tree Star (RRT*) algorithm. Firstly, RRT* is employed to generate a high-quality initial population, thereby enhancing population diversity and global search capability; secondly, the linear convergence factor is replaced with a piecewise nonlinear cosine convergence factor to balance exploration and exploitation throughout the iteration process; finally, an arctangent nonlinear inertial weight is introduced to optimise the position update mechanism, suppressing oscillations in the later stages and improving optimisation accuracy. Comparative experiments on standard test functions and 3D DEM terrain scenarios demonstrate that, compared to the best-performing benchmark algorithms in each scenario, R*WOA achieves an average improvement of 8.8% in convergence accuracy and 21.4% in optimisation stability. Its global optimisation capability significantly outperforms traditional WOA, GA and HHO algorithms, enabling the planning of optimal UAV flight trajectories with shorter paths, higher safety and better smoothness in complex constrained environments.The source code of R*WOA is publicly available at https://doi.org/10.5281/zenodo.21923510 .