Multi-Strategy Improved Particle Swarm Optimization Algorithm for Path Planning of UAV in 3-D Low Altitude Urban Environment

The Internet of Things (IoT) system and path planning algorithm provide a technological foundation for autonomous navigation of Unmanned aerial vehicles (UAVs). Geospatial data from the IoT system is transmitted to UAVs through lightweight protocols, and UAVs make optimal path decisions based on these data through optimization algorithms. The combination of the IoT, UAV, and path planning technology constitutes a UAV delivery system, which offers an efficient and economical solution for last-mile logistics in smart cities. Among these, rapid and accurate optimal path planning is crucial for the autonomous delivery of UAVs. Therefore, this article proposes a multi-strategy improved particle swarm optimization algorithm called MSIPSO. Firstly, the algorithm incorporates a local deadlock jump strategy to increase the success rate of path planning in dense obstacle environments. Secondly, to mitigate the influence of parameter selection on the algorithm’s performance, adaptive nonlinear inertia weights and learning factors are introduced to improve the algorithm’s stability. Finally, multiple population differentiation evolution strategies are designed, with different position update equations tailored for populations of varying qualities, which enhances the search efficiency of the algorithm. The simulation results show that MSIPSO outperforms particle swarm optimization (PSO), grey wolf optimizer (GWO), whale optimizer (WOA), elite archive-driven particle swarm optimization algorithm (EAPSO), and hybrid grey wolf optimizer and differential evolution algorithm (HGWODE) in terms of convergence speed, accuracy, and stability.