基于改进鹅优化算法的无人机路径规划方法研究OA

To address the problems that swarm-intelligence optimization algorithms encounter in UAV path planning namely excessive path length,insufficient path smoothness,and inadequate convergence stability,a UAV path planning method based on an improved goose optimization algorithm(GOOSE),termed GOOSE_DE,is proposed.In this method,a Sobol sequence and an opposition-based learning strategy are combined to opti-mize the initial population.A bidirectional information interaction mechanism and an intelligent selection strat-egy based on solution quality are adopted to balance global exploration and local exploitation.Furthermore,stage-wise strategy adjustment,the complementary enhancement of population diversity,and the dynamic ad-justment of the differential weight and crossover probability strengthen the algorithm's search capability and prevent it from falling into local optima.Comparative experiments are conducted against the sparrow search al-gorithm(SSA),the K-means grey wolf optimization(KMGWO)algorithm,the heterogeneous improved dynamic multi-swarm particle swarm optimization(HIDMSPSO)algorithm,the advanced-differential-evolution(ADE)al-gorithm,and the GOOSE.The results show that GOOSE_DE exhibits good convergence accuracy and stability on standard benchmark test functions.In path planning experiments on 20×20 and 30×30 grid maps,GOOSE_DE achieves the shortest average path length compared with the other five algorithms,with strong robustness and better smoothness,which fully verifies the proposed method's performance in path optimization,smoothness improvement,and robustness enhancement.