卡尔曼滤波在数学上是一种统计估算方法，通过处理一系列带有误差的实际量测数据而得到的物理参数的最佳估算。

例如在气象应用上，根据滤波的基本思想，利用前一时刻预报误差的反馈信息及时修正预报方程，以提高下一时刻预报精度。

作温度预报一般只需要连续两个月的资料即可建立方程和递推关系。

-Kalman filter in mathematics is a statistical estimation methods, By a series of errors with the actual measurement data and the physical parameters of the best estimate. For example, in meteorological applications, in accordance with the basic idea filtering, Before the use of a forecast error of the feedback information in a timely manner that prediction equation, the next moment to improve prediction accuracy. Predict the temperature usually require only two consecutive months of the establishment of equations and can be recursive relationship.