| The increased emphasis placed on safe and efficient plant operation dictates the need for advanced process control and monitoring systems that account for a host of fundamental and practical problems arising in chemical processes. Central to these problems are the issues of strong nonlinear behavior, model uncertainty, control actuator constraints and coupled discrete-continuous (hybrid) dynamics. While the co-presence of uncertainty and constraints is known to create a conflict in the control design objectives, at this stage, there is no rigorous, yet practical, approach for non-linear controller design that accounts explicitly and simultaneously for the effects of uncertainty and constraints. Furthermore, process control research has focused predominantly on continuous dynamic processes modeled by ordinary differential equations, while there are many instances when the overall process response depends on a rather intricate interaction between discrete and continuous variables. Examples of discrete events in chemical processes include inherent physico-chemical discontinuities such as phase changes and flow reversals, instrumentation with discrete settings such as binary sensors and on/off valves, logic-based switching for supervisory and safety control tasks, and actuator/sensor faults. Processes with strong discrete-continuous interactions cannot be effectively controlled with existing control techniques that do not account for the presence of discrete events.; Motivated by the above considerations, this doctoral thesis develops a general and practical framework for the analysis and control of nonlinear and hybrid process systems with uncertainty and constraints. For continuous dynamic nonlinear systems, the proposed framework yields novel Lyapunov-based nonlinear controller designs with well-characterized robustness, optimality and constraint-handling capabilities. For hybrid systems with switched dynamics, the control approach is based on the integrated synthesis, via Lyapunov techniques, of nonlinear feedback controllers with well-characterized stability regions and supervisory switching laws that orchestrate, on the basis of these regions, the transitions between the constituent modes in a way that enforces the desired control objectives in the overall closed-loop system. The proposed controller designs are applied to chemical processes of industrial interest and their effectiveness and advantages over existing control methods is documented through simulations. |
