CONTINUOUS-TIME OPTIMAL CONTROL OF A SIMULATED BOILING WATER NUCLEAR (BWR) POWER PLANT.

Abstract

A suboptimal controller has been developed for a Boiling Water Reactor Nuclear Power Plant, using the DARE P Continuous Simulation Language, which was developed in the Electrical Engineering Department at the University of Arizona. A set of 48 nonlinear first-order differential equations and a large number of algebraic equations has been linearized about the equilibrium state. Using partitioning, the linearized equations were transformed into a block triangular form. The concept of optimal control and a square performance index reflecting the desired plant behavior have been applied on the slow subsystem to develop a suboptimal controller. The obtained feedback law is shown by simulation to be able to compensate for a variety of plant disturbances. A large variety of responses can be obtained by changing the weighting matrices. The control is basically a regulator approach to speed up response during load demand changes. Several simulations are included to demonstrate the control performance. The variables to be controlled have mainly been the average neutron density and the average coolant temperature. Simplifications have been suggested, thus obtaining considerable savings in the computations and ease in design.