Model-free MIMO self-tuning controller based on support vector regression for nonlinear systems

Ucak K., Öke Günel G.

NEURAL COMPUTING & APPLICATIONS, vol.33, no.22, pp.15731-15750, 2021 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 33 Issue: 22
  • Publication Date: 2021
  • Doi Number: 10.1007/s00521-021-06194-1
  • Journal Name: NEURAL COMPUTING & APPLICATIONS
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, PASCAL, Applied Science & Technology Source, Biotechnology Research Abstracts, Compendex, Computer & Applied Sciences, Index Islamicus, INSPEC, zbMATH
  • Page Numbers: pp.15731-15750
  • Keywords: Model-free MIMO STC, STC based on SVR, Support vector regression, SVR-based parameter estimator
  • Istanbul Technical University Affiliated: Yes

Abstract

A model-free self-tuning controller (STC) based on online support vector regression (SVR) is proposed to control nonlinear and multi-input multi-output (MIMO) systems in this paper. MIMO proportional-derivative-integral (PID) controller parameters are optimized via introduced MIMO STC architecture based on SVR. The closed-loop margin notion is enhanced for MIMO type STC architectures. The adjustment mechanism is composed of only STC structure, and system model is not needed. Optimal values of STC parameters are obtained using the tracking error without any need to estimate the controlled system dynamics. In the proposed control architecture, the prediction capability of SVR and the robustness of the PID controller are combined. The success of the introduced SVR-based MIMO STC has been assessed by simulations carried out on the nonlinear Van de Vusse benchmark system. Acquired results justify that proposed structure achieves good control performance.