Moving Horizon Estimation for Continuum and Noncontinuum States with Applications in Distillation Processes

This thesis focuses on the development of advanced state estimators for continuum and noncontinuum state estimations in a switching dynamic system, and the demonstration of their applications in addressing some important process monitoring problems of distillation processes.;First, the theoretical and experimental investigations of applying a sequential continuum and noncontinuum state estimator to composition estimation in a distillation process with switching dynamics are explored. A moving horizon estimator (MHE), which has the capability to handle process constraints, is developed to estimate product composition in a distillation process under known switching mode criteria using the available temperature measurements. For situations when the system operating mode transition is unknown, an approach to state estimation under unknown switching functions is developed. The proposed method combines a MHE for composition estimation with a mode change detector to detect a change in the system operating mode and an operating mode estimator to identify the functioning mode.;Next, a noncontinuum state estimator, which is based on a moving horizon method for a class of switching system that follows a hidden Markov model, is developed. An approach to arrival cost development and constraint handling in moving horizon estimation of noncontinuum state is discussed. The effectiveness of the proposed method is illustrated by considering mode estimation problems in a simulated leakage detection system as well as a continuous stirred tank reactor.;Last, the development and application of a hybrid moving horizon estimator (HMHE) to achieve simultaneous estimation of both continuum and noncontinuum states in a constrained dynamic system is explored. One of the major issues in a moving horizon estimation approach is the development of an arrival cost to summarize the effect of past and a prior information. In this work, we have developed a generalized arrival cost, which accounts for both continuum and noncontinuum states. A generalized hybrid state estimator, which can be used as a stand-alone continuum state estimator, or as a simultaneous continuum and noncontinuum state estimator, is developed. The effectiveness of the HMHE is demonstrated through simulation studies, while its practical reliability is tested by conducting experimental studies on distillation processes.