基于PSO-LSSVM组合模型在GNSS高程拟合中的应用研究  

Research on the Application of Combined Model Based on PSO-LSSVM to GNSS Elevation Fitting

作　　者：郑建雷 肖爱龙 ZHENG Jianlei;XIAO Ailong(China Water Resources Pearl River Planning Surveying&Designing Co.,Ltd.,Guangzhou 510611,China)

机构地区：[1]中水珠江规划勘测设计有限公司,广州510611

出　　处：《广东水利水电》2025年第4期87-91,共5页Guangdong Water Resources and Hydropower

摘　　要：针对最小二乘支持向量机(least squares support vector machine,LSSVM)模型在全球导航卫星系统(global navigation satellite system,GNSS)高程拟合过程中过于依赖对参数的选择,构建PSO-LSSVM的GNSS高程拟合组合模型。通过粒子群算法(particle swarm optimization,PSO)优化最小二乘支持向量机(LSSVM)参数,采用适应度函数评估LSSVM参数粒子的质量,获取LSSSM模型最优参数,建立高程拟合模型,最后通过实例分析其拟合效果。结果表明,采用PSO-LSSVM算法构建的高程拟合模型在GNSS高程拟合中可以达到较好的拟合精度和效果,拟合精度能满足四等水准测量的精度要求。Aiming at the selection of parameters dependently of least squares support vector machine(LSSVM)in global navigation satellite system(GNSS)height fitting,the combined model based on PSO-LSSVM is established for GNSS elevation fitting.The parameters of least squares support vector machine(LSSVM)are optimized by particle swarm optimization(PSO)and the fitness function is used to evaluate the mass of each particle in order to obtain the optimal parameters of LSSVM model.Then the GNSS elevation fitting model is established and the fitting effect is analyzed by instance.The results show that the GNSS elevation fitting model based on PSO-LSSVM can achieve satisfactory precision and fitting effect and the fitting precision can meet the accuracy of fourth grade leveling survey.

关 键 词：GNSS 高程拟合 粒子群算法(PSO) 最小二乘支持向量机(LSSVM) 拟合精度 

分 类 号：P228.4[天文地球—大地测量学与测量工程]