Time Series Modeling On The Basis Of Continuous Sampling Estimation

Because the successive sampling survey can describe the dynamic change process of population of the goal,it is attracting more and more attention both here and abroad.In the known rotation scheme,we can get the target estimator of a higher precision by establishing a suitable model which could describe the true process of data generating,which has been studied maturely overseas.However,modeling in the known rotation scheme always is on the basis of a fixed rotation scheme abroad,and it is not suitable to popularize in other rotation scheme;on the other side,there is really not so many studies about time series modeling on the basis of rotation scheme at home.Basing on such a situation of sample rotation,it is very important to establish a general type model under rotation scheme.The purpose of this paper is to establish a time Series model on the basis of two way balanced one-level Rotation schemer₁^m ?r₂^m-1,then using this model to estimate the true value of study target.At last,we can use actual rotation data to check the result of the model.The purpose of this article is to give a general type time series model based on the two-dimensional balanced rotation pattern,and using the model to estimate the target population.In the estimated phase,we use the state-space model and Kalman filter to get the estimation.Then,using the current population survey,we can get the estimated results under the rotation scheme of 2⁴⁸¹ According to our national conditions,we discuss the state-space model of6³⁶² rotation scheme,and the modeling process of measurement error.At last,we look into the distance of rotation deviation and no answer question of sample rotation.It's proved that compared with other methods,the model established here is effective.