Disturbance Observer-based Subspace Predictive Control of a Pressurized Water-type Nuclear Reactor

Abstract

This work presents a disturbance observer-based predictive control strategy using a subspace matrix structure. The aim is to improve the capability of classical predictive controllers in handling external disturbances. A subspace-based predictive controller is designed directly from measurements. Then, a disturbance observer is designed using subspace matrices to estimate the external disturbance. Both of the designs are integrated using a feed-forward plus feed-back strategy to form the proposed control strategy. The proposed scheme is tested with a simulated model of a pressurized water nuclear reactor. The effectiveness of the proposed technique is demonstrated for two different load-following operations. Further, a quantitative analysis is performed to analyse the control performance of the proposed approach.