【摘要】 提出一种广义多元时变AR(autoregression)模型,并建立广义多元时变AR模型参数函数估计方法。该方法首先求得时间序列的均值函数,将广义多元时变AR模型转换为零均值多元时变AR模型,并通过谱分析和多点平均方法得到时变参数的函数形式,再分别采用最小二乘和极大似然法确定其中的待定参数。从而将一个复杂的时变问题转变为相对简单的时不变问题进行处理。该方法可广泛应用于气象、通信、自动控制、结构响应分析、故障诊断、经济分析等领域。更多还原

【Abstract】 A generalized multivariate time-varying AR(autoregression) model and a method for determining its parametric functions are presented. First,determine the mean of the time series,and then change the generalized multivariate time-varying AR model into multivariate time-varying AR model.The function forms of time-varying parameters are determined by the sample periodogram and multiple-point average.The parametric functions are obtained by least square algorithm and maximum likelihood method.Thus a complicated time-varying problem is changed into a simple time-invariant problem for further processing.The proposed method can be used in meteorology,communication,automatic control,structure response analysis,fault diagnosis,economic analysis and other fields.更多还原