Optimal tuning of fractional-order controllers based on Fibonacci-search method

Dariusz Horla; Talar Sadalla

doi:10.1016/j.isatra.2020.05.022

Optimal tuning of fractional-order controllers based on Fibonacci-search method

ISA Trans. 2020 Sep:104:287-298. doi: 10.1016/j.isatra.2020.05.022. Epub 2020 May 18.

Authors

Dariusz Horla¹, Talar Sadalla²

Affiliations

¹ Poznan University of Technology, Faculty of Control, Robotics and Electrical Engineering, Institute of Robotics and Machine Intelligence, ul. Piotrowo 3a, 60-956 Poznan, Poland. Electronic address: dariusz.horla@put.poznan.pl.
² Poznan University of Technology, Faculty of Control, Robotics and Electrical Engineering, Institute of Robotics and Machine Intelligence, ul. Piotrowo 3a, 60-956 Poznan, Poland.

PMID: 32473734
DOI: 10.1016/j.isatra.2020.05.022

Abstract

This paper presents a new optimization-based tuning method for fractional- and integer-order controllers. The method is based on minimizing an arbitrary cost function which is taken here as the integrated absolute tracking error, and it can be used for tuning purposes as well as the assessment of controller performance. The presented results are obtained from experiments with a servo drive system with a 2- and 3-parameter fractional-order controller. They depict performance improvement and ability of the proposed method to find the global minimum of a selected performance index in experimental conditions. The method does not need any model of the plant, and it has great potential for applications in general controllers or for an extension to different parameter tuning problems. The method offers a new way of changing dynamic properties of closed-loop systems, by rapid tuning of controllers. This makes possible new applications of this method, as opposed to other tuning methods, which basically require knowledge of the plant.

Keywords: Fractional-order controllers; Optimization; PI control; Performance assessment; Tuning.