All-Coefficient Adaptive Control System Design for a Space Nuclear Reactor

Abstract

Abstract Topaz-II is a space nuclear reactor developed by the Former Soviet Union. It is a thermal-neutron reactor with UO2 ceramics as fuel. Electricity is generated based on the principle of thermoelectric conversion. The reactivity is controlled by several groups control drums. Its lifetime can be three years and its reactivity feedback and further dynamic characteristics are varying with its lifetime. The positive reactivity feedback characteristics and different dynamic characteristics in different lifetime make it changeling to design a suitable control system for Topaz-II. All-coefficient adaptive control is a single-input single-output adaptive control method applied to a parameter unknown, linear time invariant or slowly varying object. In some specific conditions, it is also suitable for the non-linear systems. In this method, the model where the sum of coefficients is equal to one is reconstructed based on the original dynamic model. All coefficients in this model can be identified. Therefore, it is possible to realize the optimal control. Firstly, the model of Topaz-II nuclear reactor system is built. It includes point-reactor kinetics model with six groups delayed neutrons, a one-dimensional lumped transient thermal model of the nuclear fuel element and the moderator, a thermionic model of the converters, a control drum drive model. The reactivity feedbacks with fuel temperature and moderator are considered. Based on the model, the dynamic characteristics are analyzed. Then, with the nonlinear model of Topaz-II nuclear reactor system, the transfer functions are obtained and analyzed. Based on the transfer functions, the model where the sum of coefficients is equal to one is built by using the all-coefficient adaptive control method. After dealing with the parameter and analyzing the error, the all-coefficient adaptive control systems are designed. To evaluate the performance of the control systems, power maneuvering events including a 30%FP (Full Power) step change and a reactivity disturbance of 0.0065 are simulated. The responses with traditional and proposed control methods are evaluated and compared. The simulation results show that the all-coefficient adaptive control system has a better performance in settling time and overshoot compared with those with traditional PID controller.