地理空间信息2025，Vol.23Issue(3)：18-21,26,5.DOI:10.3969/j.issn.1672-4623.2025.03.005
改进北方苍鹰优化LSSVM的高程异常拟合算法
Improved Elevation Anomaly Fitting Algorithm of Northern Goshawk Optimization LSSVM
马亮1
作者信息
- 1. 青海省交通规划设计研究院有限公司,青海 西宁 810001
- 折叠
摘要
Abstract
Aiming at the low accuracy problem of least squares support vector machine(LSSVM)in fitting elevation anomalies,we proposed an improved the elevation anomaly fitting algorithm of Northern Goshawk optimization LSSVM.Firstly,we used chaotic mapping,adaptive inertia weight factor and Levy flight strategy to enhance the search ability of Northern Goshawk optimization algorithm,which could effectively avoid the algorithm falling into the local optimal solution.Then,we took the initial results obtained by the improved Northern Goshawk optimization algorithm as the initial parameters of LSSVM for GNSS elevation anomaly fitting,so as to increase the accuracy of elevation anomaly fitting.The results show that the algorithm can effectively improve the fitting accuracy of elevation anomaly.Compared with LSSVM and NGO-LSSVM algorithms,the external coincidence accuracy is at least 33.67%and 14.28%higher,proving the feasibility and superiority of this algorithm.关键词
改进北方苍鹰优化算法/混沌映射/自适应惯性权重因子/Levy飞行/LSSVMKey words
improved Northern Goshawk optimization algorithm/chaotic mapping/adaptive inertia weight factor/Levy flight/LSSVM分类
天文与地球科学引用本文复制引用
马亮..改进北方苍鹰优化LSSVM的高程异常拟合算法[J].地理空间信息,2025,23(3):18-21,26,5.
