Novel state-space self-tuning control for two-dimensional linear discrete-time stochastic systems

摘要

The state-space self-tuning control for 2D multi-input multi-output linear discrete-time stochastic systems is proposed in this paper, so that the output of the controlled 2D stochastic system follows (or tracks) the desired trajectory. The state-space self-tuning control methodology for the 1D stochastic systems is then extended to the 2D linear discrete-time stochastic systems. A 2D state-space self-tuning control methodology for the 2D linear discrete-time stochastic system constructs an adjustable autoregressive moving average-based noise model with estimated state first. Then, the suboptimal tracker for the 2D linear system with free boundary conditions in Roesser's model has been proposed. Based on the Roesser's model, an equivalent 1D model of the 2D system with a variable structure has been presented. More precisely, an equivalent 1D state-space innovation model is obtained in the estimating process of the 2D self-tuning control loop, and then a 2D suboptimal tracker is designed. The author 2010. Published by Oxford University Press on behalf of the Institute of Mathematics and its Applications.