Laboratory Systems Control with Adaptively Tuned Higher Order Neural Units

Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 348)

Abstract

This paper summarizes the design theory of linear and second order polynomial adaptive state-feedback controllers for SISO systems using the Batch-propagation Through Time (BPTT) learning algorithm. Deeper focus is given towards real time implementation on various laboratory experiments, with an accompaniment of corresponding theoretical simulations, to demonstrate the feasibility of use of polynomial adaptive state-feedback controllers for real time control. Raspberry Pi and open-source scripting language Python are also exhibited as a suitable implementation platform, for both testing and rapid prototyping as well as for teaching of adaptive identification and control.

Keywords

Adaptive Control Higher-Order Neural Units (HONUs) Gradient Descent (GD) Batch-Propagation Through Time (BPTT) Raspberry Pi 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Ivakhnenko, A.G.: Polynomial Theory of Complex Systems. IEEE Tran. on Systems. Man, and Cybernetics 1(4), 364–378 (1971)CrossRefMathSciNetGoogle Scholar
  2. 2.
    Taylor, J.G., Coombes, S.: Commbes, Learning higher order correlations. Neural Networks 6, 423–428 (1993)CrossRefGoogle Scholar
  3. 3.
    Kosmatopoulos, E., Polycarpou, M., Christodoulou, M., Ioannou, P.: High-Order Neural Network Structures for Identification of Dynamical Systems. IEEE Trans. on Neural Networks 6(2), 422–431 (1995)CrossRefGoogle Scholar
  4. 4.
    Nikolaev, N.Y., Iba, H.: Learning Polynomial Feed-forward Neural Network by Genetic Programming and Back-propagation. IEEE Trans. on Neural Networks 14(2), 337–350 (2003)CrossRefGoogle Scholar
  5. 5.
    Nikolaev, N.Y., Iba, H.: Adaptive Learning of Polynomial Networks: Genetic Programming, Back-propagation and Bayesian Methods. In: Genetic and Evolutionary Computation, vol. XIV, 316 p. Springer, New York (2006) ISBN: 0-387-31239-0Google Scholar
  6. 6.
    Gupta, M.M., Liang, J., Homma, N.: Static and Dynamic Neural Networks: From Fundamentals to Advanced Theory. IEEE Press and Wiley-Interscience, John Wiley & Sons, Inc. (2003)Google Scholar
  7. 7.
    Gupta, M.M., Bukovsky, I., Homma, N., Solo, M.G.A., Hou, Z.-G.: Fundamentals of Higher Order Neural Networks for Modelling and Simulation. In: Zhang, M. (ed.) Artificial Higher Order Neural Networks for Modelling and Simulation. IGI Global (2012)Google Scholar
  8. 8.
    Bukovsky, I., Homma, N., Smetana, L., Rodriguez, R., Mironovova, M., Vrana, S.: Quadratic Neural Unit is a Good Compromise between Linear Models and Neural Networks for Industrial Applications. In: The 9th IEEE International Conference on Cognitive Informatics, ICCI 2010, July 7-9. Tsinghua University, Beijing (2010)Google Scholar
  9. 9.
    Williams, R.J., Zipser, D.: A learning algorithm for continually running fully recurrent neural networks. Neural Comput. 1, 270–280 (1989)CrossRefGoogle Scholar
  10. 10.
    Williams, R.J., Peng, J.: Gradient-based learning algorithms for recurrent neural networks and their computational complexity. In: Chauvin, Y., Rumelhart, D.E. (eds.) Back-Propagation: Theory, Architectures, and Applications, pp. 433–486. Lawrence Erlbaum, Hillsdale (1992)Google Scholar
  11. 11.
    Mandic, D.P.: Generalised normalised gradient descent algorithm. IEEE Signal Process. Lett. 11, 115–118 (2004)CrossRefGoogle Scholar
  12. 12.
    Bukovsky, I., Redlapalli, S., Gupta, M.M.: Quadratic and Cubic Neural Units for Identification and Fast State Feedback Control of Unknown Non-Linear Dynamic Systems. In: Fourth Int. Symp. on Uncertainty Modelling and Analysis, ISUMA 2003, pp. 330–334. IEEE Computer Society, Maryland (2003) ISBN 0-7695-1997-0Google Scholar
  13. 13.
    Smetana, L.: Nonlinear Neuro-Controller for Automatic Control Laboratory System. Master’s Thesis, Czech Tech. Univ. in Prague (2008) (in Czech)Google Scholar
  14. 14.
    Benes, P.M.: Software Application for Adaptive Identification and Controller Tuning Student’s Conference STC, Faculty of Mechanical Engineering, CTU in Prague (2013)Google Scholar
  15. 15.
    Raspberry Pi Foundation, http://www.raspberrypi.org/about (visited January 2014)
  16. 16.
    Van Rossum, G., de Boer, J.: Linking a stub generator (AIL) to a prototyping language (Python). In: Proceedings of the Spring 1991 Europen Conference, Troms, Norway, May 20-24 (1991)Google Scholar
  17. 17.
    NumPy, package for scientific computing with Python, http://www.numpy.org/ (visited January 2014)
  18. 18.
    Hunter, J.D.: Matplotlib: A 2D Graphics Environment. Computing in Science and Engineering 9(3), 90–95 (2007), doi:10.1109/MCSE.2007.55CrossRefGoogle Scholar
  19. 19.
    Li, H., Liu, H., Gao, H., Shi, P.: Reliable Fuzzy Control for Active Suspension Systems with Actuator Delay and Fault. IEEE Transactions on Fuzzy Systems 20(2) (April 2012)Google Scholar
  20. 20.
    Hydro-pneumatic laboratory system, Automatic Control Laboratory (111), Czech Technical University in Prague, http://www1.fs.cvut.cz/cz/u12110/ar/index_c.htm (visited January 2014)
  21. 21.
    Water levitation, Automatic Control Laboratory (111), Czech Technical University in Prague, http://www1.fs.cvut.cz/cz/u12110/ar/ulohy/levitace.htm (visited January 2014)
  22. 22.
    Sláma, M.: Using Minicomputer Raspberry Pi for Adaptive Identification and Control of Hydro-pneumatic Laboratory System with Python. Diploma Thesis, FS, ČVUT in Prague, defended (January 2014) (in Czech)Google Scholar
  23. 23.
    Bukovský, I., Rodriguez, R., Bíla, J., Homma, N.: Prospects of Gradient Methods for Nonlinear Control. Strojárstvo Extra 5(5), art. no. 36, 1–5 (2012) ISSN 1335-2938Google Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  1. 1.Department of Instrumentation and Control EngineeringCzech Technical University in PraguePragueCzech Republic

Personalised recommendations