A hybrid improved Grey Wolf optimization algorithm for three dimensional UAV path planning in complex terrain
TL;DR
Experiments show that HLGWO generally outperforms several comparison algorithms in convergence accuracy, stability, and path cost, thereby improving the safety, feasibility, and optimization performance of 3D UAV path planning in complex environments.
Abstract
To address the tendency to fall into local optima, insufficient convergence accuracy, and path-quality fluctuations in three-dimensional UAV path planning under complex terrain and multiple constraints, this study proposes a hybrid improved Grey Wolf Optimization algorithm, termed HLGWO. A unified objective function is first constructed by considering path length, safety risk, flight altitude, turning smoothness, and terrain complexity, and an adaptive weighting mechanism is introduced to meet the requirements of different flight stages. Within the standard GWO framework, Latin Hypercube Sampling is used to improve the initial population distribution, Gaussian random walk is incorporated to enhance local search capability, and a Differential Evolution operator is introduced to promote information exchange and refined exploitation among individuals. Experiments on the CEC2005 and CEC2020 benchmark suites, together with eight real DEM-based UAV flight scenarios, show that HLGWO generally outperforms several comparison algorithms in convergence accuracy, stability, and path cost, thereby improving the safety, feasibility, and optimization performance of 3D UAV path planning in complex environments.