TABLE OF CONTENTS Model predictive control has been proven by a group of companies to provide benefits, in the right applications, that are greater than those achieved from the improvement of basic control systems [9.17]. The greatest benefits are realized in applications with dead-time dominance, interactions, constraints, and the need for some optimization. The constraints define the operating limits of processes or equipment. Optimization is often as simple as the maximization or minimization of a flow. The advantage of model predictive control lies in its knowledge of the effects of past actions of manipulated and disturbance variables on the future profile of controlled and constraint variables.
For systems with large dead times, interactions, and multiple constraints, the ability to provide the necessary patience, anticipation, and forecast of the future approach to targets and limits is essential for moving an operation closer to its optimum. "A good control engineer can draw straight lines; a great one can move the lines." Model predictive control is a tool that can make a good control engineer a great one.
Many good review papers [9.1], [9.2], [9.3] and reports [9.4], [9.5] have been published on model predictive control (MPC). In contrast, there are only a few books dedicated to the subject [9.11], [9.12]. The literature published to date focuses on the theory or on an overview of the application and benefits of MPC. Information on how to design, build, commission, and tune an MPC controller has been lacking.
