方程(1)称为过程模型(或状态空间方程);方程(2)称为观测模型。式中X_(k,k-1)为系统状态的r阶向量,即由第k-1步预报第k步的过程模型值;X_(k-1)为第k-1步的过程模型值:U_(k-1)为输入向量,即第k-1步的上游或支流的输入值;W_(k-1)为第k-1步的模型噪声,即第k-1步的模型误差:Φ_(k-1)、_(k-1)、Γ_(k-1)、Η_k与G_k为参数向量,说明系统动态性质,Z_k为第k步观测值。 Equation (1) is called the process model (or state space equation); equation (2) is called the observation model. Where X_ (k, k-1) is the r-order vector of the system state, that is, the process model value of the k-th step is predicted by the k-1 step; and X k-1 is the process model value of the k- : U_ (k-1) is the input vector, that is, the input value of upstream or tributary of k-1 step; W k-1 is the model noise of k- : Φ_ (k-1), _ (k-1), Γ_ (k-1), Η_k and G_k are parameter vectors, which describe the dynamic properties of the system. Zk is the observation value of the kth step.