Transition control techniques in nonlinear process control

In this thesis, a local modeling/controller approach is presented for the transition control problem in chemical processes. A multi-model predictive control algorithm is developed for systems represented by a sequence of piecewise linear models along the transition trajectory. The control algorithm is a receding horizon scheme with a quasi-infinite horizon objective function that has finite and infinite horizon cost components. The finite horizon cost consists of free input variables that direct the system towards a terminal region that contains the desired operating point. The infinite horizon cost has an upper bound and takes the system to the final operating point. The control problem is formulated as a convex optimization in terms of Linear Matrix Inequalities. A quantitative criterion for recognizing which model is in effect and for deciding the sequence of models is also included in the control structure. The control strategy is enhanced by the incorporation of output feedback by the design of an observer for systems represented by multiple models. The performance of the proposed multi-model predictive control is illustrated on a CSTR and an industrial scale solution copolymerization reactor.; The control strategy is further modified by the addition of a contractive constraint to guarantee stability. The use of piecewise linear models and a switching criterion has added a hybrid characteristic to the control structure. Therefore, a recent approach in analysis of hybrid systems called multiple Lyapunov functions is utilized in the stability analysis of the closed-loop system.