Application of a model-reference adaptive controller to a space nuclear power system

Abstract

This paper presents the results of applying a model-reference adaptive controller with selective state-variable weighting to a nonlinear simulation of an SP-100-type space nuclear power system. The results demonstrate the utility and safety benefit of utilizing a model-reference adaptive controller in a space nuclear power system that experiences a change in its fuel temperature reactivity feedback coefficient. The change in the fuel temperature reactivity feedback coefficient produces an undesirable system response to an open loop insertion of external reactivity that is made to respond to an increase in electrical power demand by the payload. Results for both step and ramp reactivity input signals show that selective state-variable weighting incorporated with a modelreference adaptive controller can successfully cause the plant state variable of choice to track the analogous reference-model state variable response to a reactivity input during the transient portion of the system response; this is demonstrated for the nuclear reactor thermal power, the nuclear fuel temperature, and the temperature of the thermoelectric power conversion assembly.