Software-based optimal PID design with robustness and noise sensitivity constraints. (September 2015)

• Record Type: Journal Article
• Authors: Garpinger, Olof; Hägglund, Tore
• Is Part Of: Journal of process control. Volume 33(2015:Sep.)
• Page Start: 90
• Page End: 101
• Publication Date: 2015-09
• Subjects: Control system design -- Measurement noise -- Optimization -- PID control -- Software tools; Process control -- Periodicals; Fabrication -- Contrôle -- Périodiques; Process control; Periodicals; Electronic journals; 660.281
• DOI: 10.1016/j.jprocont.2015.06.001
• Languages: English
• ISSNs: 0959-1524

Abstract

Abstract : Highlights: Proposes a new software-based method for optimal PID and filter design with robustness and noise sensitivity constraints. The design procedure selects the most suitable controller from a set of PID, PI and I controllers. Performance, robustness and noise sensitivity trade-offs are visualized for three representable process types. Presents a new method for determining how much different process types benefit from derivative action. Abstract: Even though PID control has been available for a long time, there are still no tuning methods including derivative action that have gained wide acceptance in industry. Also, there is still no general consensus for when one should use PID, PI or even I control on a process. The focus of this article is to present a new method for optimal PID control design that automatically picks the best controller type for the process at hand. The proposed PID design procedure uses a software-based method to find controllers with optimal or near optimal load disturbance response subject to robustness and noise sensitivity constraints. It is shown that the optimal controller type depends on maximum allowed noise sensitivity as well as process dynamics. The design procedure thus results in a set of PID, PI and I controllers with different noise filters that the user can switch between to reach an acceptable control signal activity. The software is also used to compare PI and PID control performance with equivalent noise sensitivity and robustness over a large batch of processes representative for the process industry. This can be used to show how much a particular process benefits from using the derivative part.