Adaptive one-step-ahead optimal controller based on WLS schemes

An optimal adaptive control technique for the discrete linear systems is discussed in this paper. The system parameters are unknown and one-step-ahead adaptive control design is based on the input matching approach and the weighted least-squares (WLS) algorithm. It is shown that the adaptive stochastic system is globally closed-loop stable and the system identification is consistent. The adaptive controller converges to the one-step-ahead optimal controller. Finally, some simulation examples are given to demonstrate the reliability of the new optimal adaptive control algorithm. Copyright © 2008 John Wiley & Sons, Ltd.