Accuracy Comparison between F-RLS algorithm for CARAR systems and F-RLS algorithm for CARARMA systems

In this paper, the data filtering based recursive least squares algorithm for a CARAR systems, and data filtering based recursive least squares algorithm for a CARARMA systems are derived for comparison. These algorithms are based on the decomposition technique and in this technique, the main algorithm transform into two sub algorithms with smaller sizes. First, System identification model and another is the noise identification model. The problem here is the unknown variables in the information vectors and the used idea for solving this problem is to replacing these unknown variables with their corresponding estimates. Thus, the parameters of these two identification models can be estimated using recursive least squares method. Finally, a simulation example is provided to support the comparison between these proposed algorithms.