A Novel Hybrid Enhanced Particle Swarm Optimization for UAV Path Planning
Jiehong Wu, Youzheng Sun, Jing Bi, Chengcheng Chen, Yuanmang Xie
Abstract
With the application of uncrewed aerial vehicles(UAVs) in rescue and logistics delivery, safe and feasible path planning in dense environments becomes particularly important. Responding to the challenge, this paper proposes a hybrid enhanced particle swarm optimization (PSO) algorithm, namely GEPSO. The algorithm first enhances the PSO algorithm with three strategies to improve its convergence speed and adaptability and then combines it with the gravitational search algorithm (GSA). In this process, the acceleration of GSA is used as a new component for velocity update in the enhanced PSO, thereby strengthening the optimization ability of the algorithm and effectively balancing the capabilities of exploitation and exploration. In addition, this study employs cubic spline interpolation to fit the flight path, making the path smoother and more in line with the flight characteristics of UAVs. The experiments are simulated in a three-dimensional environment, compared with algorithms such as the PSO, GSA, spherical vector particle swarm optimization (SPSO), and differential evolution cylindrical vector particle swarm optimization (DEPSO), and the results show that the GEPSO algorithm can plan paths with higher stability and better quality in complex three-dimensional environments.