PI controller for SISO linear systems based on neural linear PCA

In this paper an approach to design proportional-integral (PI) controllers, for SISO systems, based on neural linear principal components analysis (PCA) is presented. Closed-loop control can be formulated and implemented within the reduced space defined by a PCA model. The neural linear PCA controller, results in an integral controller, which can be used as an inferential controller. The main contributions of the paper are: a) the proposed architecture with a classical proportional controller and a neural integral controller based on linear neural PCA; b) the evaluation of the controller performance using the Harris index. Some experimental results obtained with a DC motor linear model are presented, showing the controller performance.