摘要
Thereexistsagreatdealofperiodicnon-stationaryprocessesinnatural,socialandeco-nomicalphenomenon.Itisveryimportanttorealizethedynamicanalysisandreal-timeforecastwithinaperiod.Inthisletter,awavelet-Kalmanhybridestimationandforecastingalgorithmbasedonstep-by-stepfilteringwiththereal-timeandrecursionpropertyisputforward.ItcombinestheadvantagesofKalmanfilterandwavelettransform.Utilizingtheinformationprovidedbymulti-sensoreffectively,thisalgorithmcanrealizenotonlyreal-timetrackinganddynamicmulti-stepfore-castingwithinaperiod,butalsothedynamicforecastingbetweenperiods,andithasagreatvaluetothesystemdecision-making.Simulationresultsshowthatthisalgorithmisvaluable.
出版日期
2007年05月15日(中国Betway体育网页登陆平台首次上网日期,不代表论文的发表时间)