Doctor of Philosophy
Electrical and Computer Engineering
Dr. Jin Jiang
The dynamic characteristics of Canadian Supercritical Water-cooled Reactor (SCWR) are significantly different from those of CANDU reactors due to the supercritical water coolant and the once-through direct cycle coolant system. Therefore, it is necessary to study its dynamic behaviour and further design adequate control system.
An accurate dynamic model is needed to describe the dynamic behaviour. Moving boundary method is applied to improve numerical accuracy and stability. In the model construction process, three regions have been considered depending on bulk and wall temperature being higher or lower than the pseudo-critical temperature. Benefits of adopting moving boundary method are illustrated in comparison with the fixed boundary method. The model is validated with both steady-state and transient simulation and can accurately predict the dynamic behaviour of the Canadian SCWR.
A linear dynamic model, for dynamic analysis and control system design, is obtained through linearization on the nonlinear dynamic models derived from conservation equations. The linearized dynamic models are validated against the full order nonlinear models in both time domain and frequency domain. The open-loop dynamics are also investigated through extensive simulations.
Cross-coupling analysis among inputs and outputs is examined using Relative Gain Array (RGA) and Nyquist plots, from which adequate input-output pairings are identified. Cross-coupling at different operating conditions are also evaluated to illustrate the nonlinearities. It can be concluded that the Canadian SCWR is a Multiple Input and Multiple Output (MIMO) system with strong cross-coupling and a high degree of nonlinearity.
Due to the existence of strong cross-coupling, the Direct Nyquist Array (DNA) method is used to decouple the system into a diagonal dominance form via a pre-compensator. Three Single Input and Single Output (SISO) compensators are synthesized to the pre-compensated system in the frequency domain. The temperature variation induced by the disturbances at the reactor power and pressure can be significantly reduced. To deal with the nonlinearities, a gain scheduling control strategy is adopted. Different set of controllers are used at different load conditions. The control strategy is evaluated under various operating scenarios. It is shown that gain scheduling control can successfully achieve satisfactory performance for different operating conditions.
Sun, Peiwei, "Dynamic Model Construction and Control System Design for Canadian Supercritical Water-cooled Reactors" (2012). Electronic Thesis and Dissertation Repository. 896.