复杂环境下基于改进混合蜣螂优化算法的无人机三维路径规划方法  

Improved hybrid dung beetle algorithm for UAV 3D path planning in complex environments

作　　者：姜鹏洲 张琳 程钦 赵耀 JIANG Pengzhou;ZHANG Lin;CHENG Qin;ZHAO Yao(School of Electrical and Information Engineering,Jiangsu University of Technology,Changzhou 213000,China;SunwaveCommunications Co.,Ltd.,Hangzhou 320053,China;School of Communications and Information Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210023,China)

机构地区：[1]江苏理工学院电气信息工程学院,江苏常州213000 [2]三维通信股份有限公司,杭州320053 [3]南京邮电大学通信与信息工程学院,南京210023

出　　处：《兵器装备工程学报》2025年第3期203-215,共13页Journal of Ordnance Equipment Engineering

基　　金：国家自然科学基金项目(62341119);江苏省基础研究计划(青年基金)项目(BK20210941);常州市领军型创新人才引进培育项目(CQ20210094)。

摘　　要：针对复杂环境下无人机路径优化算法收敛精度低、全局搜索能力弱及易陷入局部最优解的问题,提出了一种改进混合蜣螂优化算法(SPM and osprey based hybrid dung beetle optimizer,SO-DBO)。使用混沌映射SPM初始化种群位置,提高算法搜索效率。在滚球蜣螂种群有障碍模式和无障碍模式中分别引入动态全局勘探策略和随机角度策略,提升算法精度和全局搜索能力。在觅食蜣螂位置更新引入自适应T分布策略,增强算法逃离局部最优能力。通过动态权重因子提高算法全局搜索能力并降低陷入局部最优解的风险。实验结果表明:相比原始蜣螂优化算法(dung beetle optimizer,DBO)和粒子群算法(particle swarm optimization,PSO),改进后的SO-DBO算法代价函数指标在简单环境下分别改善了9.68%、12.93%,在复杂环境下分别改善了13.34%、17.00%,有效提升了算法的收敛速度、精度和稳定性。To address the difficulty of low convergence accuracy,weak global search ability,and susceptibility to getting stuck in local optima of unmanned aerial vehicle(UAV)path optimization algorithms in complex obstacle environments,a new improved hybrid dung beetle optimization algorithm(SPM and Osprey based Hybrid Dung Beetle Optimizer,SO-DBO)combines chaotic mapping and Osprey optimization is proposed.The Sine Piece Wise Linear Chaotic Map(SPM)is used to initialize the population positions which improves the search efficiency of the algorithm.This study introduces dynamic global exploration strategy and Gaussian distributed random angle strategy in both obstacle free and obstacle free modes of the rolling ball beetle population to improve algorithm accuracy and global search capability.Introducing an adaptive T-distribution strategy to update the position of foraging beetles enhances the algorithm’s ability to escape local optima.By customizing dynamic weight factors,the algorithm’s global search capability is improved and the risk of getting stuck in local optima is reduced.The experimental study shows that compared to the original Dung Beetle Optimizer(DBO)and Particle Swarm Optimization(PSO),the SO-DBO algorithm improves the cost function index by 9.68%and 12.93%in simple environment,and 13.34%and 17.00%in complex environment,respectively,which effectively improves the convergence speed,accuracy and stability of the algorithm.

关 键 词：复杂环境 蜣螂优化算法 无人机路径规划 混沌映射 代价函数 

分 类 号：V249[航空宇航科学与技术—飞行器设计]