论文部分内容阅读
针对目前土壤湿度反演方法研究较少且缺少实时性的现状,该文提出一种土壤湿度反演方法——最小二乘支持向量机技术。以积分方程模型为正向算法,数值模拟不同雷达参数(频率、入射角及极化)下后向散射系数随土壤含水量和地表粗糙度的变化情况。经过数据敏感性分析,选取C-波段和X-波段、小入射角下的同极化后向散射系数作为支持向量回归的训练样本信息;经过适当的训练,利用支持向量回归技术对土壤含水量进行了反演研究;并考虑通过多频率、多极化、多入射角数据的组合,消除地表粗糙度的影响,提高反演精度。模拟结果表明,该方法反演土壤湿度具有较高的精度和较好的实时性;同时,与人工神经网络方法的结果比较,证明了该方法的有效性,为土壤湿度的反演研究提供了一种方法。
Aiming at the current situation of less research on soil moisture inversion and its lack of real-time performance, this paper presents a soil moisture inversion method-least squares support vector machine. Taking the integral equation model as the forward algorithm, the numerical simulation of the change of the backscattering coefficient with the soil water content and surface roughness under different radar parameters (frequency, angle of incidence and polarization) was carried out. After data sensitivity analysis, the co-ordinated backscattering coefficient at C-band and X-band and the co-polarized backscattering coefficient at small incident angle were chosen as training sample information for regression of support vector. After appropriate training, the support vector regression The inversion is studied. The combination of multi-frequency, multi-polarization and multi-angle data is also considered to eliminate the influence of surface roughness and improve the inversion accuracy. The simulation results show that this method has higher accuracy and better real-time performance than the artificial neural network method, and the validity of the proposed method is proved. This method can provide the inversion of soil moisture a way.