Decoupling control of a nuclear reactor by using a dynamic compensator

In a large-sized nuclear reactor the space dependency of the dynamic behaviour becomes outstanding, and a more advanced control method than usual region control is desired. The decoupling theory completed by Gilbert cannot be applied directly to the reactor, because a state variable (delayed neutron precursor) is not measurable. In this paper, applying the decoupling theory after introducing the dynamic compensator, the reactor system is decoupled only by the measurable state variables, and favorable dynamic characteristics are sought for. Some numerical examples are given. A comparison between the obtained results and that obtained based on the adaptation of the observer is also given. (auth.)