Modeling and model updating in the real-time optimization of gasoline blending

This thesis provides a new approach to gasoline blender control that can out-perform conventional blend controllers. The proposed blend controller includes appropriate nonlinear blending models and adopts concepts from model predictive control theory in order to allow it to handle stochastic disturbances in feedstock qualities. The resulting Real-Time Optimization (RTO) system is similar to model predictive control in that it predicts disturbances over one time horizon and optimizes the blender control problem over another. It then implements control action for the current time-step and repeats the process in a receding horizon fashion.;Another important contribution is that of parameter observability for steady-state RTO systems. The currently available method provides only a necessary but not sufficient condition. A new approach has been presented which uses fundamental statistical principles and is applicable to any steady-state system where secondary measurements are used to estimate unmeasured quantities. Observability is then extended to a measure of degree of observability.