Data-driven Koopman fractional order PID control of a MEMS gyroscope using bat algorithm

Mehran Rahmani, Sangram Redkar

Research output: Contribution to journalArticlepeer-review

Abstract

Data-driven control methods are strong tools due to their predictions for controlling the systems with a nonlinear dynamic model. In this paper, the Koopman operator is used to linearize the nonlinear dynamic model. Generating the Koopman operator is the most important part of using the Koopman theory. Dynamic mode decomposition (DMD) is used to obtain eigenfunction for producing the Koopman operator. Then, a fractional order PID (FOPID) controller is applied to control the linearized dynamic model. A swarm intelligence bat optimization algorithm is utilized to tune the FOPID controller’s parameters. Simulation results on micro-electromechanical systems (MEMS) gyroscope under conventional PID controller, FOPID, Koopman-based FOPID controller (Koopman-FOPID), and Koopman-FOPID control optimized by bat algorithm (Koopman-BAFOPID) show that the proposed Koopman-BAFOPID controller has better performance in comparison with three other controllers in terms of high tracking performance, low tracking error, and low control efforts.

Original languageEnglish (US)
JournalNeural Computing and Applications
DOIs
StateAccepted/In press - 2023

Keywords

  • Bat algorithm
  • Data-driven method
  • Dynamic mode decomposition
  • Fractional PID control
  • Koopman operator
  • MEMS gyroscope

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'Data-driven Koopman fractional order PID control of a MEMS gyroscope using bat algorithm'. Together they form a unique fingerprint.

Cite this