A Method of Optimization of Flux and Xenon Fluctuations in a Nuclear Reactor

The problem of controlling the spatial flux distribution in a reactor is formulated as an optimization problem. A nodal representation of the distributed parameter system that results in a set of non-linear ordinary differential equations is used. This non-linear optimal control problem is solved directly by Differential Dynamic Programming. Several numerical examples are presented to show the efficiency of the method, the short computational time needed for convergence, and the wide range of sudden demands that can be handled.