卡尔曼滤波回归统计模型及工程应用分析

将统计模型的回归系数看做状态向量，统计模型视为观测方程，利用卡尔曼滤波算法实现了统计模型的建模。该建模方法不仅便于实时处理、计算机实现、节省机时，而且由于卡尔曼滤波属于线性最小方差估计，所以相比最小二乘建模方法可望获得更高的建模精度。工程实践也验证了卡尔曼滤波回归统计模型的上述特性。

Study on Kalman filtering Regression Statistic model and Projects Application Analysis

In this paper, the regression coefficient of the statistical model was looked as the state vector, the statistical model was as observation equations, and then it using Kalman filtering algorithm to achieve the statistical modeling, the modeling method is not only convenient for real-time processing, computer realization, saving time, but because Kalman filtering is belong to the linear minimum square estimation, it is expected to achieve the higher modeling precision compared to least square method. Project has proved the view of Kalman filtering statistical regression models.