In this paper, the recursive optinal instrumental vairable method, which estimates noise dynamics as well as system dynamics using pre-filtered data, was propsed. In case of no pre-filtering, the computational burden of this algorithm is equal to that...
In this paper, the recursive optinal instrumental vairable method, which estimates noise dynamics as well as system dynamics using pre-filtered data, was propsed. In case of no pre-filtering, the computational burden of this algorithm is equal to that of extended least squares(ELS) methods and less than that of the rescursive maximum likelihood(RML) method. But in case of prefiltering the computational burden of this algorithm is equal to that of the RML, but this algorithm shows faster convergence than ELS and RML algoorithms.
Also, vaeious instrumental variables were proposed, and various types of prefilters were prposed to improve the performance of the proposed algorithms.
Computer simulations were performed to illustrate the efficiency of the proposed methods.