全球定位系统2026，Vol.51Issue(3)：27-33,7.DOI:10.12265/j.gnss.2026052
基于梯度优化器融合策略的快速选星方法
A fast satellite selection method gradient-based optimizer fusion strategy
韩艾航 1戴卫恒 2袁普钊 1吕晶 1李广侠1
作者信息
- 1. 南京航空航天大学 电子信息工程学院,南京 211106
- 2. 陆军工程大学 通信工程学院,南京 210007
- 折叠
摘要
Abstract
To meet the requirements of"high precision,low time consumption and strong robustness"of satellite selection algorithms for satellite positioning in complex environments,a fast satellite selection method based on the fusion strategy of gradient-based optimizer(GBO)is proposed.This method integrates the GBO algorithm with the Levy flight mechanism combined with Cauchy mutation and Rice mutation that is multi-mutation gradient-based optimizer(MMGBO)method,forming a multi-mode mutation mechanism.It effectively enhances the global exploration ability and local escape ability of the GBO algorithm,while alleviating the stagnation problem in the later stage of convergence.Simulation experiments show that the positioning accuracy of the MMGBO algorithm is superior to that of the single GBO algorithm,and the GDOP error compared with the traversal method is generally close to zero.The calculation efficiency improvement ratio can even reach more than 99%.The proposed method provides an efficient and reliable satellite selection scheme for real-time high-precision positioning of GNSS combined with low-Earth orbit satellites,and can be adapted to multi-system fusion positioning scenarios such as BeiDou Navigation Satellite System(BDS)and GPS.关键词
选星/基于梯度的优化器/几何精度衰减因子(GDOP)/Levy飞行Key words
satellite selection/gradient-based optimizer/geometric dilution of precision(GDOP)/Levy flight分类
天文与地球科学引用本文复制引用
韩艾航,戴卫恒,袁普钊,吕晶,李广侠..基于梯度优化器融合策略的快速选星方法[J].全球定位系统,2026,51(3):27-33,7.
