一种适用于电离层层析成像的TV-MART算法

A TV-MART Algorithm Applied to Computerized Ionospheric Tomography

摘要: 针对电离层电子密度重构问题,提出了一种综合利用总变差最小化(toalvariation,TV)与乘法代数重构算法(multiplicativealgebraicreconstructiontechnique,MART)的电离层层析成像算法。该算法对反演模型的参数施加总变差约束,以提高反演过程的稳定性和结果的精确性。通过模拟数据和实测数据对模型的重构结果进行验证,结果表明,相对于乘法代数重构算法,该算法能够有效地提高电离层电子密度的重构精度。

Abstract: Reconstructionofionosphericelectrondensity(IED)isanill posedinverseproblem.Tosolvethisproblem,weproposeanewalgorithmcombiningtotalvariation(TV)minimizationwithamultiplicativealgebraicreconstructiontechnique(MART)forcomputerizedionospherictomography(CIT).ThealgorithmisaninterlacediterativemethodaddingaTVconstrainttotheparameteroftheinversionmodel,toimprovethestabilityandtheresolutionofinversion.Experimentswithsimulatedandrealdataarediscussed.IncontrasttotheMART,theproposedalgorithmcaneffectivelyimproveIEDreconstructionaccuracy.