The Kalman filtering of 2-D systems in Fornasini-Marchesini model

In this paper, the traditional Kalman filtering problem of two- dimensional systems is addressed. Different from the existing results which possess special constraints on filter structure, we propose an estimator without any structure constraint such that an exactly optimal solution to the minimization of the variance of estimation error can be achieved. Similar to the existing 1-D Kalman filter, the proposed 2-D Kalman filter can also be computed recursively.