Robust optimal control of nuclear reactors

This work presents an approach for incorporating advanced control algorithms in power plants. The approach uses existing conventional and accepted controllers and accomplishes an advanced control objective by augmenting the demand signals to conventional feedback control loops, State Feedback Assisted Classical Control (SFAC). Optimal control theory is used to design an SFAC controller to provide tight control of reactor temperatures while reactor power regulation is provided by a classical output feedback controller. A robust controller configuration is obtained using a Luenberger Observer modified for the known nonlinear characteristic of a reactor. The gain of the embedded classical controller can also be used to tune the robustness characteristics of the system.; The robustness of the controller is assessed using (1) linear analysis of a combined model of a higher order plant and low order controller and examining the sensitivity of the combined model's dominant eigenvalues to wide-range plant parameter variations and (2) extensive nonlinear simulation with higher order plant models which included the Babcock and Wilcox Modular Modeling System reactor model for a Three Mile Island type Pressurized Water Reactor. A fixed nonlinear time invariant controller with good reactor temperature performance is obtained for the major yearly plant parameter variations due to fuel burnup and control rod movement in the power range 10 to 100 percent.; A real world power plant is nonlinear, stochastic, and subject to some uncertainties. However, it is a central premise of this work that linear control systems analysis techniques can be utilized to design robust control of such processes which accommodate uncertainties between the real world plant and mathematical modeling for control systems analysis and synthesis. The modeling and analysis presented in this work should immediately provide a platform to develop, test and demonstrate alternate approaches to robust reactor and power plant control design.