关键词: 无人机航迹规划, 粒子群算法, 雷达传感器, 自适应惯性权重, 柯西-高斯变异

Abstract: A Cauchy Gauss Particle Swarm Optimization (CGPSO) algorithm based on CGPSO is proposed to address the slow convergence and the tendency to fall into local optimum in traditional PSO UAV trajectory planning algorithms for forestry resource protection tasks. The UAV mission environment model is constructed by pre-screening the forest environment with radar sensors; adaptive inertia weights and fused Cauchy-Gauss variational operators are introduced to adjust the particle swarm algorithm to balance the global-local convergence speed and optimize the local extreme problem; the UAV track length cost, obstacle collision cost and elevation range cost are comprehensively analyzed and a track planning fitness function is built. Simulation results show the standard deviation of the planned algorithm’s fitness reaches 0.148 6 and the time taken is 54.34 s, which is 42% less than the convergence generation value of the PSO algorithm and 25% better than the time taken. It is feasible to use the new planning trajectory algorithm for forestry resource protection in forest areas.