摘要
时序模型的参数估计是时序模型建模的核心。当数据长度增加到一定值后 ,传统的递推最小二乘法会出现数据饱和现象 ,使模型参数远离真值。为此 ,本文在分析用于传统时序模型参数估计的递推最小二乘法基础上 ,给出了模糊加权递推最小二乘法的基本理论、递推算式和加权方法 。
Parameter estimation is the key in building up a model in time series analysis Datum saturation will emerge and the parameters estimated will depart from actual ones in conventional recurrence least square algorithm when the data length increases beyond a threshold value The paper presents the basic theory and recurrence formulas for a fuzzy weighted recurrence least square algorithm in the time series model based on an analysis of the conventional recurrence least square algorithm The results of application of the simulation and analysis are applied upon random road roughness for military engineering vehicles
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2001年第1期120-123,共4页
Acta Armamentarii
