摘要
Order-recursiveleast-squares(ORLS)algorithmsareappliedtotheprob-lemsofestimationandidentificationofFIRorARMAsystemparameterswhereafixedsetofinputsignalsamplesisavailableandthedesiredorderoftheunderlyingmodelisunknown.Onthebasisofseveraluniversalformulaeforupdatingnonsymmetricprojec-tionoperators,thispaperpresentsthreekindsofLSalgorithms,callednonsymmetric,symmetricandsquarerootnormalizedfastORLSalgorithms,respectively.Astotheau-thors’knowledge,thefirstandthethirdhavenotbeensofarprovided,andthesecondisoneofthosewhichhavethelowestcomputationalrequirement.Severalsimplifiedversionsofthealgorithmsarealsoconsidered.
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)