摘要
本文针对上次对角双线性时序模型(简记为USDBL(p,q,r))进行研究.在输入为正态白噪声的假定下,通过广义传递函数得到了该模型平稳性的充分必要条件,并计算了输出序列的谱密度函数.证明了该模型的二阶特性与一个线性平稳ARMA模型相似,最后给出了该模型的最优化线性预报方法.
In this paper, we discuss upper Sub-Diagonal Bilinear Time Series Models (USDBL models). By the general transfer functions, stationary conditions and spectral desity function are obtained for USDBL models. It is shown that the second order structure is similar to a linear ARMA model with uncorrelated errors. In the end, the best linear predictors are given for USDBL models.
出处
《东南大学学报（自然科学版）》
EI
CAS
CSCD
1993年第3期63-72,共10页
Journal of Southeast University：Natural Science Edition
关键词
谱密度
线性预报
双线性
时序模型
bilinear, transfer functions, spectral density linear predictors / stationarity
