Recursive Least Squares Algorithm for MISO CARAMA Systems Utilizing Data Filtering

The identification problem of multiple-input single-output (MISO) output-error for (CARARMA) systems is concerned in this paper. the filtering technique is used to decompose the main model into two identification models, one for system and another for noise, and the difficulty is the information vector has unknown variables. The employed idea for solving this difficulty is replacing these variables with their corresponding estimates, and a data filtering-based RLS method is derived. The proposed algorithm has more significant computational advantage and can estimate parameters more accurately compared with the RELS algorithms. This conclusion is confirmed in simulation results.