Model Reference Robust Adaptive Control of Control Element Drive Mechanism in a Nuclear Power Plant

Abstract

In the nuclear power plant, the control element drive mechanism (CEDM: A type of solenoid actuators) is used to insert nuclear fuel rods into the core to regulate the neutron absorption within the reactor. The control environment is full of uncertainty and disturbance, and the CEDM is supposed to cope with the varying conditions in the reactor. In this paper, a model reference adaptive control (MRAC) of CEDM is investigated. The existing MRAC scheme that assumes the known system order is first extended to the system with unknown high-frequency gain and output disturbance, and then the extended version is applied to the control problem of CEDM. A simplified relationship between the input voltage and the magnetic force of the CEDM is established based on Kirchhoff’s voltage law and the magnetic co-energy theory. The system order of the plant is identified by the prediction error method using experimental input/output data: A 3rd order linear model with three poles and two zeros turns out to be the best model for control. Based on this model, a 3rd order reference model is selected, and an adaptive control law with s-modified adaptation laws is designed under output disturbance. The performance of the proposed control law is compared with a well-tuned PI controller and the conventional MRAC. The superiority of the proposed method is also demonstrated in an abnormal condition. Simulation and experiment results are provided.