Considerations in the Design and Implementation of Control Laws for the Digital Operation of Research Reactors

Abstract

Factors relevant to the design and implementation of digital controllers for research reactors are discussed with emphasis on the rationale for incorporating a system model in the control law. For this purpose, proportional-integral-derivative and period-generated control are compared. The latter is a model-based technique that achieves excellent trajectory tracking of nonlinear systems. It does this by combining feedback and feedforward control action in a manner that cancels the effects of the system’s dynamics on the controller’s performance. Model-based control is also superior in that it permits replication of some of the functions that humans perform when exercising control. In particular, models can be used to predict expected plant response and thereby facilitate diagnosis. The importance of validated signals, supervisory algorithms, properly designed man-machine interfaces, and automated diagnostics are discussed in relation to control law implementation. In addition, a summary is provided of reactor dynamics as related to control, and arguments are presented in support of using the rate of change of reactivity as the actuator signal. Experimental results obtained from trials of digital controllers on both the 5-MW(thermal) Massachusetts Institute of Technology Research Reactor and the Annular Core Research Reactor that is operated by Sandia National Laboratories are included.