Global control of the furuta pendulum using artificial neural networks and feedback of state variables

  • Luisa F. Escobar-Dávila Universidad Tecnológica de Pereira, Pereira
  • Oscar D. Montoya-Giraldo Universidad Tecnológica de Pereira, Pereira
  • Didier Giraldo-Buitrago Universidad Tecnológica de Pereira, Pereira
Keywords: Mathematical model, Furuta Pendulum, state variables, artificial neural networks, hybrid control.

Abstract

This paper presents the mathematical modeling of the Furuta Pendu-lum by power functions, taking into account the non linear own dynamics of the physical systems and considering the existing couplings between the electric and mechanic devices. A control process based on feedback of state variables (FSV) for the equilibrium point is developed and two topics for the non linear zone are addressed. First of all, functions are implemented to represent the energetic states of the plant in a global way and the operation regions are established (“Swing up” zone), and later Artificial Neural Networks (ANN) are employed to simulate the behavior of the energy functions. Finally, it is presented the combination between the control techniques, considering the own constrains of the actuators and sensors used, besides of this, a study is done in a simulated environment of the physical phenomena that may disturb system behavior, and the capacity, sensitivity and robustness of the controller is verified.

Author Biographies

Luisa F. Escobar-Dávila, Universidad Tecnológica de Pereira, Pereira
Ing. electricista, Facultad de Ingeniería,
Universidad Tecnológica de Pereira, Pereira
Oscar D. Montoya-Giraldo, Universidad Tecnológica de Pereira, Pereira
Ing. electricista, Facultad de Ingeniería,
Universidad Tecnológica de Pereira, Pereira
Didier Giraldo-Buitrago, Universidad Tecnológica de Pereira, Pereira
Ing. electricista, Facultad de Ingeniería,
Universidad Tecnológica de Pereira, Pereira
How to Cite
Escobar-Dávila, L. F., Montoya-Giraldo, O. D., & Giraldo-Buitrago, D. (2013). Global control of the furuta pendulum using artificial neural networks and feedback of state variables. TecnoLógicas, (30), 71-94. https://doi.org/10.22430/22565337.89

Downloads

Download data is not yet available.
Published
2013-06-30
Section
Research Papers

Most read articles by the same author(s)