基于天鹰算法优化下的FastSLAM2.0算法  

A FastSLAM2.0 Algorithm Based on Aquila Optimizer

作　　者：员鑫涛 张炳琪 

机构地区：[1]长安大学,陕西西安710064

出　　处：《大众科技》2023年第2期16-20,59,共6页Popular Science & Technology

摘　　要：针对FastSLAM2.0的重采样过程因频繁重采样而出现粒子退化,导致建图精度降低的现象,将FastSLAM2.0算法与天鹰算法相结合,提出一种基于天鹰算法优化下的FastSLAM2.0算法,以提高建图精度和改善粒子退化现象。在算法中通过天鹰算法优化粒子寻优策略,同时对重采样过程中粒子权重较小的粒子进行交叉、变异操作,增大粒子多样性,缓解粒子退化现象,提高机器人位姿估计一致性。在基于ROS平台下的实体样机上对改进的算法进行可靠性验证。实验结果表明:改进的优化算法能有效提高定位建图精度。In view of the phenomenon that the resampling process of FastSLAM2.0 results in particle degradation due to frequent and heavy sampling,thus reducing the accuracy of mapping,the FastSLAM2.0 algorithm is combined with the aquila optimizer,and a FastSLAM2.0 algorithm based on the aquila optimizer optimization is proposed to improve the accuracy of mapping and particle degradation,.In the algorithm,aquila optimizer is used to optimize the particle optimization strategy.At the same time,crossover and mutation operations are carried out for particles with small particle weight in the resampling process to increase particle diversity alleviate particle degradation,and improve the consistency of robot pose estimation.The reliability of the improved algorithm is verified on a solid prototype based on ROS platform.The experimental results show that the improved optimization algorithm can effectively improve the accuracy of location mapping.

关 键 词：天鹰优化算法 FASTSLAM 遗传重采样 ROS 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]