Estimation and Identification for Correlation Coefficient Stationary Series

The estimation method for the parameters of the correlation coefficient stationary series which can not be measured straightly is presented.After the state equations of the correlation coefficient AR(p),MA(q) and ARMA(p,q)series are established,the filtering,prediction and smoothing for non stationary series can be carried out.In addition,the identification method of the observation system is also given.For the relation between time and the mean,the variance of the noise can be considered,the precision of the present method is higher than that of the traditional methods.