An IMC-Based Extended PID Controller for Sampled-Data Systems

Extended Proportional Integral Derivative (EPID) is a model based control algorithm which consists of a conventional digital PID controller plus an additional compensator. This novel extension to the PID control algorithm offers the same properties as the IMC controller for a broad class of common industrial process models. EPID control simplifies the industrial implementation of IMC and has three important additional advantages. First, since EPID is based on the PID algorithm, it is easily understood by process engineers, control engineers and maintenance personnel. Second, the EPID strategy is easy to implement on industrial distributed control systems and can take full advantage of the functionality built into the available PID algorithms such as set-point tracking, automatic mode switching, etc. Third, since the IMC design procedure for sampled-data systems is used to derive the EPID controller parameters, it results in a controller that eliminates intersample rippling of the manipulated variable and overshoot of the process variable. Additionally, EPID explicitly takes into account model uncertainty, and has only one adjustable parameter which is directly related to the closed-loop speed of response and to control robustness. This paper presents the derivation of the EPID algorithm and IMC-based EPID tuning rules for common industrial process models. The practical and powerful EPID design tools, including performance analysis and robust stability analysis allow the designer to achieve the best compromise between the conflicting objectives of performance and robustness.