Control of nonlinear processes operating at various steady states: Analysis and synthesis of linear and adaptive model based control approaches

Control of nonlinear systems has always been the focus of process industry, simply for the fact that many processes such as high purity distillation columns, highly exothermic chemical reactions, pH neutralizations, and batch systems can exhibit "highly" nonlinear response. If the process is only "mildly" nonlinear or remains in the 'vicinity' of a nominal steady state, then the effects of nonlinearity may not be severe. However these processes may be required to operate over a wide range of conditions due to "large" process upsets or setpoint changes. When conventional PID controllers are used to control "highly" nonlinear processes, the controllers must be tuned very conservatively in order to provide stable behavior over the entire range of operating conditions. However conservative controller tuning can result in serious degradation of control system performance.;Thus, there are considerable incentives for developing more effective control strategies that incorporates knowledge of the nonlinear process. This incentive combined with continuous improvement in the capabilities of computer-control hardware and software, making it feasible to incorporate complex nonlinear models in plant control systems, encouraged a big interest in theory and applications of nonlinear control. However this interest lacks a smooth transition from linear control to nonlinear control applications. The questions when to use nonlinear control or when and which linear control action will be sufficient for the control of a nonlinear process in case nonlinearities are "mild", must be answered for the proper application of nonlinear control and linear control techniques.;In this work, a methodology for the analysis and synthesis of model based linear controllers is developed, for sufficient linear control of a nonlinear process operating at different steady states. For the cases where linear control would not achieve desired performance, a nonlinear control technique, simultaneous model predictive control and identification (MPCI) of previous studies, is considered. MPCI, an adaptive MPC methodology, which uses convenient linear models of the process at various steady states, found by closed-loop identification by on-line optimization, can be an alternative to nonlinear MPC. However closed-loop theoretical properties of this technique, are missing. In this work, those theoretical properties of MPCI are established.