论文部分内容阅读
提出了一种支持向量回归机(SVRM)辅助的北斗地球静止轨道(GEO)卫星反射信号土壤湿度反演方法。使用全球导航卫星系统反射信号(GNSS-R)右旋圆极化(RHCP)天线和左旋圆极化(LHCP)天线接收体制进行了地基实验,采集了北斗GEO卫星直射、反射信号原始数据,并从中提取直射、反射信号的相关功率,结合北斗GEO卫星的高度角与方位角信息作为输入,烘干称重法获取的土壤湿度作为输出对使用径向基(RBF)核函数的ε-SVRM进行了训练。独立测试集上的结果表明,SVRM辅助的北斗GEO卫星反射信号土壤湿度反演方法获取的土壤湿度结果与烘干称重法获取的土壤湿度参考值误差控制在3%以内,线性回归方程决定系数为0.8979,均方根误差RMSE为1.4926%,证明了该方法具有良好的泛化特性,实际应用中效果良好。
A method of soil moisture retrieval from Beidou Geostationary Earth Orbit (GEO) satellite reflectance signal, which is supported by Support Vector Regression (SVRM), is proposed. Ground tests were conducted using the GNSS-R RHCP antenna and the LHCP antenna receiving system. The raw data of the direct and reflected signals of the Beidou GEO satellite were collected and analyzed. The relative power of direct and reflected signals was extracted. Based on the altitude and azimuth information of Beidou GEO satellite as input, the soil moisture obtained by drying and weighing method was used as output to conduct ε-SVRM using radial basis function (RBF) kernel function Training. The results on the independent test set show that the error of soil moisture obtained by SVRM-assisted soil moisture inversion method of GEO satellite reflected signal and the reference value of soil moisture obtained by the dry weighing method is controlled within 3%, and the linear regression equation determines the coefficient Is 0.8979, and RMSE of root mean square error is 1.4926%. It proves that this method has good generalization characteristics and has good effect in practical application.