Unmanned Aerial Vehicle 3D Path Planning Based on Improved Nonlinear Particle Swarm Optimization Algorithm

When executing a mission, an Unmanned Aerial Vehicle (UAV) must rely on three-dimensional path planning to avoid obstacles. Particle Swarm Optimization (PSO) is suitable for the path planning problem because of few parameters, fast convergence and easy implementation. However, along with the dynamic change of the search path, the particle diversity tends to be singularity, and also the population easily fall into the local optimal region. In this regard, An Improved Nonlinear Particle Swarm Optimization (INPSO) Algorithm is proposed for UAV 3D path planning. This algorithm introduces the Sugeno function to construct nonlinear inertia weights and learning factors for enhancing the global search capability of the particle population. At the same time, INPSO incorporates a reverse learning elimination mechanism based on the wolf predation strategy to update the poorly adapted particles in the late iteration, so as to regulate the particle population diversity and enhance the optimal seeking and convergence performance of the algorithm. Simulation tests show that compared with other algorithms, INPSO not only achieves the shortest average path, but also obtains a higher effective path rate in complex terrain environments, and has better convergence speed and search accuracy.