Robust optimal control algorithms for magnetic levitation plant

M. V. Sotnikova, Yu A. Gilyazova, E. A. Selitskaya

Research output: Contribution to journalConference articlepeer-review


Robust control design is one of the main branches of research activity in modern control theory. This is determined by special practical importance of the issues discussed here and the necessity to create new effective methods and algorithms for robust control design based on modern capabilities of digital computers. The main goal of robust control design is to provide some desired robust properties of the closed loop system, such as stability and quality of control processes, while the mathematical model can vary within given admissible limits. In the paper the problem of robust control design for magnetic levitation plant is considered. This system has an inaccurate mathematical model, because it has essentially nonlinear dynamics and very difficult approximate description of the agnetic field. Two different optimization approaches for control design are proposed. The first of them is based on the use of predictive models and the second one is the modification of a linear-quadratic synthesis, where additional constraints in the frequency domain are introduced. The corresponding computational algorithms for real-time implementation of a control laws are given. These algorithms are compared by illustrative examples of magnetic levitation system control.

Original languageEnglish
Pages (from-to)325-334
Number of pages10
JournalCEUR Workshop Proceedings
StatePublished - 1 Jan 2017
Event2nd International Scientific Conference "Convergent Cognitive Information Technologies", Convergent 2017 - Moscow, Russian Federation
Duration: 24 Nov 201726 Nov 2017

Scopus subject areas

  • Computer Science(all)


  • Control algorithms
  • Magnetic levitation
  • Optimization
  • Predictive model
  • Real-time implementation
  • Robust control

Fingerprint Dive into the research topics of 'Robust optimal control algorithms for magnetic levitation plant'. Together they form a unique fingerprint.

Cite this