Title :
Iterative instrumental variable method for robust identification of systems
Author :
Dai, Heping ; Sinha, Naresh K.
Author_Institution :
Atlantic Nucl. Services Ltd., Fredericton, NB, Canada
fDate :
10/1/1995 12:00:00 AM
Abstract :
A new off-line method, called the robust iterative instrumental variable method with modified residuals, is presented for robust identification of systems in this paper. It is based on Huber´s minimax principle and the instrumental variable principle. Proof of convergence for the proposed method is given along with some results of simulation. Comparison with both nonrobust methods and robust methods proposed earlier shows that the proposed method provides better accuracy of estimate and speed of convergence
Keywords :
convergence of numerical methods; estimation theory; iterative methods; minimax techniques; parameter estimation; simulation; Huber´s minimax principle; convergence; instrumental variable principle; iterative instrumental variable method; modified residuals; nonrobust methods; off-line method; robust identification; simulation; systems identification; Convergence; Cost function; Instruments; Iterative algorithms; Iterative methods; Least squares methods; Minimax techniques; Niobium; Noise robustness; Parameter estimation;
Journal_Title :
Industrial Electronics, IEEE Transactions on
Conference_Location :
10/1/1995 12:00:00 AM