A generalized stochastic adaptive control algorithm

Le résumé fourni par la source

A generalized stochastic state space adaptive control algorithm which combines system identification based on minimum variance, Newton, and gradient search approaches, with an original control law derived from a per-interval performance index. This unique approach results in superior accuracy and performance. The identification process (prediction error approach) is based on the state space innovations model, which eliminates the use of the Kalman filtering algorithm for the state estimation, therefore is computationally efficient. The control law for tracking is generated from a per-interval (one-step ahead prediction) performance index, is shown to have higher performance and computationally more efficient than that using the dynamic programming approach.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.