Least squares learning identification

This paper presents learning identification method for a class of stochastic systems with time-varying parametric uncertainties. The least squares learning algorithm is derived on the basis of repetitive operations over a pre-specified finite time interval. The repetitive persistent excitation is shown to be a sufficient condition for establishing convergence of the learning algorithm. It is shown that the estimates converge to the time-varying values of the parameters over the entire interval, and the complete estimation is achieved. The effectiveness of the proposed learning algorithm is demonstrated by the given numerical results.