used to filter the latent state variable-we used to filter

卡尔曼滤波以最小均方误差为最佳估计准则，采用信号与噪声的状态空间模型，利用前一时刻的估计值和当前时刻的观测值来更新对状态变量的估计，求出当前时刻的估计值，算法根据建立的系统方程和观测方程对需要处理的信号做出满足最小均方误差的估计-Kalman filter to minimize the mean square error criterion for the best estimates, using the state space model of signal and noise, usin