Fast error whitening algorithms for system identification and control with noisy data

signal (MSE) being WC) to e noise. rmined ecursive hm has ed. One . In the es this ed cost ithm to lts with oblems. Linear system identification with noisy input/output is a critical problem in processing and control. Conventional techniques based on the mean squared-error criterion can at best provide a biased parameter estimate of the unknown system modeled. Recently, we proposed a new criterion called the error whitening criterion (E solve the problem of linear parameter estimation in the presence of additive whit Accordingly, the central idea is to partially whiten the error signal beyond a predete correlation lag. In the first-half of the paper, we will derive a fast Quasi-Newton type r algorithm to compute the optimal EWC solution in an online manner. The algorit O(N) complexity where, N represents the length of the parameter vector to be estimat of the primary limitations of EWC is the assumption that the input noise must be white second-half of this paper, we will introduce a modified cost function that overcom assumption and allows the noise in the input to be colored. The analysis of this modifi function is then presented followed by a sample-by-sample stochastic gradient algor optimally compute the analytical solution. Finally, we will show the experimental resu EWC as well as the modified criterion in system identification and controller design pr r 2005 Elsevier B.V. All rights reserved.