于胜杰, 万蓉, 付志康. 代数重构算法在GNSS水汽层析解算中的应用[J]. 武汉大学学报 ( 信息科学版), 2016, 41(8): 1113-1117, 1124. DOI: 10.13203/j.whugis20140316
引用本文: 于胜杰, 万蓉, 付志康. 代数重构算法在GNSS水汽层析解算中的应用[J]. 武汉大学学报 ( 信息科学版), 2016, 41(8): 1113-1117, 1124. DOI: 10.13203/j.whugis20140316
YU Shengjie, WAN Rong, FU Zhikang. Application of Algebraic Reconstruction Technique on the GNSS Water Vapor Tomography[J]. Geomatics and Information Science of Wuhan University, 2016, 41(8): 1113-1117, 1124. DOI: 10.13203/j.whugis20140316
Citation: YU Shengjie, WAN Rong, FU Zhikang. Application of Algebraic Reconstruction Technique on the GNSS Water Vapor Tomography[J]. Geomatics and Information Science of Wuhan University, 2016, 41(8): 1113-1117, 1124. DOI: 10.13203/j.whugis20140316

代数重构算法在GNSS水汽层析解算中的应用

Application of Algebraic Reconstruction Technique on the GNSS Water Vapor Tomography

  • 摘要: 全球导航卫星系统(GNSS)水汽层析反演技术是目前获取对流层水汽三维分布的重要方法。考虑到代数重构算法在迭代反演中具有节省计算机内存且计算稳定度高的优点,对代数重构算法在GNSS水汽层析中的应用进行了研究。研究结果表明,受水汽在对流层中的分布情况的影响,传统的加法代数重构算法在实际的层析解算中,会出现较大的重构误差,而乘法代数重构算法和调整了松弛参数向量的加法代数重构算法则大大提高了层析解算的精度;代数重构算法较附加约束条件的层析解算方法更易受到观测值误差的影响,但采用乘法代数重构算法可以获得优于加法代数重构算法的结果。

     

    Abstract: The GNSS water vapor tomography technique can be used to obtain spatially resolved humidity information about the troposphere. The application of this method in GNSS water vapor tomography is discussed in detail considering the need to save computer memory and in light of the high stability when calculating inversion with the algebraic reconstruction technique, The Algebraic Reconstruction Technique is used to construct the distribution of water vapor; experimental results indicate that the solution of the traditional ART shows large reconstruction error due to the distribution property of the water vapor in the troposphere, while the IART method which adapts a relaxation parameter vector gets a favorable solution. The MART method also shows the similar results. Compared inversion with constraints, the algebraic reconstruction technique method is more susceptible to the observation error. The solution generated by the MART method is better than that of the IART method.

     

/

返回文章
返回