| The wastewater chlorination process in fixed-size open-channel reactors is a challenging control problem because of the presence of a variable dead time caused by changes in the inlet flow rate. The approach proposed here involves the utilization of a moving weir as an actuator in a novel residence time control scheme, designed to ensure that fluid passing through the reactor experiences a constant residence time independent of flow rate. The dynamic weir control loop is conceived to supplement existing control structures that are charged with manipulating the dosage of chlorine. Simulations studies show that both feedforward and feedback manifestations of the residence time controller are highly effective at improving chlorine dosage control and are highly robust with respect to weir-flow modeling and measurement errors. In fact, these controllers have the potential to improve the control performance dramatically while saving as much as 45% of the chlorine used for treatment.; An automatic control system described as run-to-run is one in which a discrete controller is applied to an inherently discrete process, and where the performance measurements are made only after all processing is completed. A complete theoretical analysis is developed for the case of a static plant under proportional-integral control, resulting in quantitative descriptions of the nominal and robust stability regions, and recommended tuning heuristics. The analysis for the case of a Smith Predictor design shows that very small modeling errors in the estimated measurement delay may cause closed-loop instability. Therefore, the use of a Smith Predictor for run-to-run systems suffering from uncertainty in the delay is not recommended.; A weak-acid/strong-base neutralization control system utilizing two alternative pH sensors is modeled and analyzed. The first sensor is a traditional pH electrode whose output signal is proportional to pH, and the second sensor is a fiber-optic optrode that produces a signal that is proportional to hydrogen ion concentration. Systematic simulation studies show that no inherent performance improvement is observed when using the optrode instead of the electrode sensor in the feedback loop. |
