基于卡尔曼滤波估计的连续性抽样调查研究

A Study of Successive Sampling Survey on the Basis of Kalman Filter Estimation

针对连续性抽样调查中如何提高连续调查数据准确性的问题,本文引入时间序列分析方法,分别考虑连续性抽样调查中的重复样本和轮换样本等不同情况,建立了连续性抽样调查下的状态空间模型,利用成熟的卡尔曼滤波估计方法给出了总体均值的估计量。由于状态空间模型及卡尔曼滤波估计方法能够充分利用各期连续样本的调查信息,给出了精度更高的估计量,从而能够产生更加准确的连续性时间序列数据。

In order to improve the precision of successive data under successive sampling survey, this paper uses the time series analysis. Respectively considering repeated sample and rotation sample, state-space model under successive survey is constructed and the estimators of population mean are presented through Kalman filter methods. These estimators would be more precise because more bypassed information can be used through the state-space model and Kalman filter estimation. Therefore, the precision of successive time series data would be improved further more.