Account for uncertainty with robust control design: Part 1

The dynamic behavior of a process can never be modeled perfectly. Process controllers can be designed to handle this uncertainty. The main goal of control theory is the design of a controller that senses the behavior of the process, compares it to the desired behavior, computes corrective actions based on a model of the system, and prompts the system to take the corrective action. Most controller design methods require a linear nominal model of the process. Some techniques rely on frequency-domain analysis, although more recent methods can be used to formulate time- and frequency-domain specifications as optimization problems.