Abstract
An original approach to the estimation of transfer functions of linear discrete time or continuous time systems from noisy input and output measurements is presented. The proposed frequency domain Maximum Likelihood Estimator (MLE) assumes independent Gaussian noise on both the input and output Fourier coefficients. It is demonstrated that most of the properties of the Gaussian MLE remain unchanged when it is applied to measured input and output Fourier coefficients corrupted by non-Gaussian errors. From this a robust Gaussian frequency domain estimator results, which is very useful for the practical identification of linear time invariant systems. The theoretical results are verified by simulations and experiments.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
