NEW EFFICIENT ORDER-RECURSIVE LEAST-SQUARES ALGORITHMS

在线阅读 下载PDF 导出详情
摘要 Order-recursiveleast-squares(ORLS)algorithmsareappliedtotheprob-lemsofestimationandidentificationofFIRorARMAsystemparameterswhereafixedsetofinputsignalsamplesisavailableandthedesiredorderoftheunderlyingmodelisunknown.Onthebasisofseveraluniversalformulaeforupdatingnonsymmetricprojec-tionoperators,thispaperpresentsthreekindsofLSalgorithms,callednonsymmetric,symmetricandsquarerootnormalizedfastORLSalgorithms,respectively.Astotheau-thors’knowledge,thefirstandthethirdhavenotbeensofarprovided,andthesecondisoneofthosewhichhavethelowestcomputationalrequirement.Severalsimplifiedversionsofthealgorithmsarealsoconsidered.
机构地区 不详
出版日期 1989年02月12日(中国期刊网平台首次上网日期,不代表论文的发表时间)