Set-theoretic estimation based on a priori knowledge of the noise distribution

A new algorithm far estimation of a linear-in parameters model is developed and tested by simulation. The method is based on the assumption of indepen dent, identically distributed noise samples with a triangular density funct ion. Such a noise model well approximates the symmetrically distributed sou rces of noise frequently encountered in practice, and the inclusion of a di stribution assumption allows the computation of a pseudo-mean estimate to c omplement the set solution. The proposed algorithm recursively incorporates incoming observations with decreasing computational complexity as the numb er of updates increases. Simulations demonstrate that the algorithm Las ver y favorable convergence rates and estimation accuracy and is very robust to deviations from the assumed noise properties. Comparisons with other set-t heoretic algorithms and with conventional RLS are given.