Abstract:
In this paper, a Pressurized Heavy Water Reactor (PHWR) is modeled in discrete domain. Based on model predictive control (MPC) technique, three separate models are identified for reactor power, logarithmic reactor power and rate logarithmic reactor power. The objectives of the proposed reference racking multivariable composite model predictive controller are to minimize the difference between desired and predicted linear reactor power, desired and predicted logarithmic reactor power and desired and predicted rate logarithmic reactor power and the variation of moderator level in the nuclear reactor core. A controlled auto-regressive and integrated moving average (CARIMA) model is used for mathematical modeling of controller design. Constraints are imposed in control and rate of control of moderator level in the nuclear reactor core. A multi-objective constrained optimization problem is solved for composite model predictive controller design using MATLAB. The performance of the proposed controller is tested for a reactor start-up condition of a PHWR-type nuclear power plant operating inPakistanand found excellent and fast in reference tracking.
Page(s):
245-252
DOI:
DOI not available
Published:
Journal: Proceedings of Pakistan Academy of Sciences, Volume: 47, Issue: 4, Year: 2010