RLS based estimation schemes for self-tuning control

Abstract
An adaptive or a self-tuning controller is a combination of an on-line, recursive estimation law with a suitable control strategy. The performance and convergence properties play an important role in the overall stability and performance of an adaptive controller. Recursive least squares (RLS) is probably the most popular parametric identification method in adaptive control. RLS is one member of a family of prediction error identification methods that are based on the minimization of prediction error functions. The basic RLS method is the result of minimization of a quadratic cost function of the prediction error. This chapter provides a brief introduction to the basic RLS algorithm and its convergence properties, and then surveys the various modifications of this algorithm.

This publication has 0 references indexed in Scilit: