Robust Least Square Method and Its Application to Parameter Estimation

The system identification problem is researched when the input and output signal are both corrupted by noise. The robust least square (RLS) method and its application to parameter estimation problem, in which the perturbations are unknown but bounded (UBB), are introduced. The method can be interpreted as Tikhonov regularization procedure, with the advantage that it provides an exact bound on the robustness of solution and a rigorous way to compute the regularization parameter. Simulation results verify that the estimation precision and the robustness anti-noise of RLS are remarkably higher than other method when the input and output signal are both corrupted by noise.