Title :
Recursive least squares algorithm with compensation for coloured noise
Author_Institution :
Dept. of Electr. & Electron. Eng., Victoria Univ. of Technol., Footscray, Vic., Australia
Abstract :
A novel algorithm based on RLS (recursive least squares) is proposed for identifying the system dynamics in the presence of colored noise. The expected biased estimation from RLS is suitably compensated. Theoretical analysis and simulation studies show that the algorithm should and does deliver convergent estimations without SPR (strictly positive real) or like conditions as normally imposed in ELS (extended least squares) schemes
Keywords :
identification; noise; statistical analysis; coloured noise compensation; identification; recursive least squares; strictly positive real condition; system dynamics; Algorithm design and analysis; Analytical models; Colored noise; Convergence; Feedback; Least squares approximation; Least squares methods; Linear regression; Linear systems; Recursive estimation; Resonance light scattering; Stability;
Conference_Titel :
Decision and Control, 1991., Proceedings of the 30th IEEE Conference on
Conference_Location :
Brighton
Print_ISBN :
0-7803-0450-0
DOI :
10.1109/CDC.1991.261693
