A family of model predictive control algorithms with artificial neural networks

Maciej Ławryńczuk

International Journal of Applied Mathematics and Computer Science (2007)

  • Volume: 17, Issue: 2, page 217-232
  • ISSN: 1641-876X

Abstract

top
This paper details nonlinear Model-based Predictive Control (MPC) algorithms for MIMO processes modelled by means of neural networks of a feedforward structure. Two general MPC techniques are considered: the one with Nonlinear Optimisation (MPC-NO) and the one with Nonlinear Prediction and Linearisation (MPC-NPL). In the first case a nonlinear optimisation problem is solved in real time on-line. In order to reduce the computational burden, in the second case a neural model of the process is used on-line to determine local linearisation and a nonlinear free trajectory. Single-point and multi-point linearisation methods are discussed. The MPC-NPL structure is far more reliable and less computationally demanding in comparison with the MPC-NO one because it solves a quadratic programming problem, which can be done efficiently within a foreseeable time frame. At the same time, closed-loop performance of both algorithm classes is similar. Finally, a hybrid MPC algorithm with Nonlinear Prediction, Linearisation and Nonlinear optimisation (MPC-NPL-NO) is discussed.

How to cite

top

Ławryńczuk, Maciej. "A family of model predictive control algorithms with artificial neural networks." International Journal of Applied Mathematics and Computer Science 17.2 (2007): 217-232. <http://eudml.org/doc/207833>.

@article{Ławryńczuk2007,
abstract = {This paper details nonlinear Model-based Predictive Control (MPC) algorithms for MIMO processes modelled by means of neural networks of a feedforward structure. Two general MPC techniques are considered: the one with Nonlinear Optimisation (MPC-NO) and the one with Nonlinear Prediction and Linearisation (MPC-NPL). In the first case a nonlinear optimisation problem is solved in real time on-line. In order to reduce the computational burden, in the second case a neural model of the process is used on-line to determine local linearisation and a nonlinear free trajectory. Single-point and multi-point linearisation methods are discussed. The MPC-NPL structure is far more reliable and less computationally demanding in comparison with the MPC-NO one because it solves a quadratic programming problem, which can be done efficiently within a foreseeable time frame. At the same time, closed-loop performance of both algorithm classes is similar. Finally, a hybrid MPC algorithm with Nonlinear Prediction, Linearisation and Nonlinear optimisation (MPC-NPL-NO) is discussed.},
author = {Ławryńczuk, Maciej},
journal = {International Journal of Applied Mathematics and Computer Science},
keywords = {optimisation; neural networks; quadratic programming; linearisation; predictive control},
language = {eng},
number = {2},
pages = {217-232},
title = {A family of model predictive control algorithms with artificial neural networks},
url = {http://eudml.org/doc/207833},
volume = {17},
year = {2007},
}

TY - JOUR
AU - Ławryńczuk, Maciej
TI - A family of model predictive control algorithms with artificial neural networks
JO - International Journal of Applied Mathematics and Computer Science
PY - 2007
VL - 17
IS - 2
SP - 217
EP - 232
AB - This paper details nonlinear Model-based Predictive Control (MPC) algorithms for MIMO processes modelled by means of neural networks of a feedforward structure. Two general MPC techniques are considered: the one with Nonlinear Optimisation (MPC-NO) and the one with Nonlinear Prediction and Linearisation (MPC-NPL). In the first case a nonlinear optimisation problem is solved in real time on-line. In order to reduce the computational burden, in the second case a neural model of the process is used on-line to determine local linearisation and a nonlinear free trajectory. Single-point and multi-point linearisation methods are discussed. The MPC-NPL structure is far more reliable and less computationally demanding in comparison with the MPC-NO one because it solves a quadratic programming problem, which can be done efficiently within a foreseeable time frame. At the same time, closed-loop performance of both algorithm classes is similar. Finally, a hybrid MPC algorithm with Nonlinear Prediction, Linearisation and Nonlinear optimisation (MPC-NPL-NO) is discussed.
LA - eng
KW - optimisation; neural networks; quadratic programming; linearisation; predictive control
UR - http://eudml.org/doc/207833
ER -

References

top
  1. Kesson B. M. and Toivonen H. T. (2006): A neural network model predictive controller. - J. Process Contr., Vol.16, No.3, pp.937-946. 
  2. Bacic M., Cannon M. and Kouvaritakis B. (2002): Feedback linearization MPC for discrete-time bilinear systems. - Proc. 15-th IFAC World Congress, Barcelona, Spain, CD-ROM, paper 2391. 
  3. Babuška R., Sousa J. M. and Verbruggen H. B.(1999): Predictive control of nonlinear systems based on fuzzy and neural models. - Proc. European Control Conf., Karlsruhe, Germany, CD-ROM, paper F1032-5. 
  4. Bazaraa M. S., Sherali J. and Shetty K. (1993): Nonlinear Programming: Theory and Algorithms. - New York: Wiley. Zbl0774.90075
  5. Bloemen H.H.J., van den Boom T. J. J. and Verbruggen H. B. (2001): Model-based predictive control for Hammerstein-Wiener systems. - Int. J. Contr., Vol.74, No.5, pp.482-495. Zbl1015.93022
  6. Brdyś M.A. and Tatjewski P. (2005): Iterative algorithms for multilayer optimizing control. - London: Imperial CollegePress/World Scientific. Zbl1083.93001
  7. Cavagnari L., Magni L. and Scattolini R. (1999): Neural network implementation of nonlinear receding-horizon control.- Neural Comput. Applic., Vol.8, No.1, pp.86-92. 
  8. Clarke D. W., Mohtadi C. and Tuffs P. S. (1987): Generalized predictive control - I. The basic algorithm. - Automatica, Vol.23, No.2, pp.137-148. Zbl0621.93032
  9. Cutler R. and Ramaker B. (1979): Dynamic matrix control - A computer control algorithm. - Proc. AIChE National Meeting, Houston. 
  10. Dutka A. and Ordys A. W. (2004): The optimalnon-linear generalised predictive control by the time-varying approximation.- Proc. 10-th IEEE Int. Conf. Methods and Models in Automation and Robotics, Międzyzdroje, Poland, pp.299-303. 
  11. Grimble M.J. and Ordys A.W. (2001): Nonlinear predictive control for manufacturing and robotic applications. - Proc. 7-th IEEE Int. Conf. Methods and Models in Automation and Robotics, Międzyzdroje, Poland, pp.579-592. 
  12. Haykin S. (1999): Neural Networks - A Comprehensive Foundation. - Englewood Cliffs, NJ: Prentice Hall. Zbl0934.68076
  13. Henson M. A. (1998): Nonlinear model predictive control: Current status and future directions. - Comput. Chemi. Engi., Vol.23, No.2, pp.187-202. 
  14. Hornik K., Stinchcombe M. and White H. (1989): Multilayer feed forward networks are universal approximators. - Neural Netw., Vol.2, No.5, pp.359-366. 
  15. Hussain M. A. (1999): Review of theapplications of neural networks in chemical process control - Simulation andonline implementation. - Artifi. Intelli. Eng., Vol.13, No.1, pp.55-68. 
  16. Kavsek B.K., Skrjanc I. and Matko D. (1997): Fuzzy predictive control of a highly nonlinear pH process. - Comput. Chem. Eng., Vol.21, Supplement, pp.S613-S618. 
  17. Kouvaritakis B., Cannon M. and Rosser J. A.(1999): Nonlinear model based predictive control. - Int.J. Contr., Vol.72, No.10, pp.919-928. Zbl0984.93038
  18. Liu G. P., Kadirkamanathan V. and Billings S. A. (1998): Predictive control for non-linear systems using neural networks. - Int. J. Contr., Vol.71, No.6, pp.1119-1132. Zbl0948.90053
  19. Ławryńczuk M. and Tatjewski P. (2006): An efficient nonlinear predictive control algorithm with neural models and its application to a high-purity distillation process. - Lecture Notes in Artificial Intelligence, Springer, Vol.4029, pp.76-85. 
  20. Ławryńczuk M. and Tatjewski P. (2004): A stable dual-mode type nonlinear predictive control algorithm basedon on-line linearisation and quadratic programming. - Proc.10-th IEEE Int. Conf. Methods and Models in Automationand Robotics, Międzyzdroje, Poland, pp.503-510. 
  21. Ławryńczuk M. (2003): Nonlinear model predictive control algorithms with neural models. - Ph.D. thesis, Warsaw University of Technology, Warsaw, Poland. 
  22. Ławryńczuk M. and Tatjewski P. (2003): An iterative nonlinear predictive control algorithm based on linearisation and neural models. - Proc. European Control Conf., Cambridge, U.K., CD-ROM, paper 339. Zbl1151.93342
  23. Ławryńczuk M. and Tatjewski P. (2002): A computationally efficient nonlinear predictive control algorithm based on neural models. - Proc. 8-th IEEE Int. Conf. Methods and Models in Automation and Robotics, Szczecin, Poland, pp.781-786. 
  24. Ławryńczuk M. and Tatjewski P. (2001): A multivariable neural predictive control algorithm. - Proc. IFAC Advanced Fuzzy-Neural Control Workshop, Valencia, Spain, pp.191-196. 
  25. Maciejowski J.M. (2002): Predictive Control with Constraints. - Harlow, U.K.: Prentice Hall. 
  26. Mahfouf M. and Linkens D.A. (1998): Non-linear generalized predictive control (NLGPC) applied to muscle relaxant anaesthesia. - Int. J. Contr., Vol.71, No.2, pp.239-257. Zbl0938.93527
  27. Maner B.R., Doyle F.J., Ogunnaike B.A. and Pearson R.K. (1996): Nonlinear model predictive control of a simulated multivariable polymerization reactor using second-order Volterra models. - Automatica, Vol.32, No.9, pp.1285-1301. Zbl0875.93324
  28. Michalska H. and Mayne D.Q. (1993): Robust receding horizon control of constrained nonlinear systems. - IEEE Trans. Automat. Cont., Vol.38, No.11, pp.1623-1633. Zbl0790.93038
  29. Morari M. and Lee J. (1999): Model predictive control: Past, present and future. - Comput. Chem. Engi.,Vol.23, No.4/5, pp.667-682. 
  30. Nørgaard M., Ravn O., Poulsen N. K. and Hansen L.K. (2000): Neural Networks for Modelling and Control of Dynamic Systems. - London: Springer. Zbl0953.93003
  31. Osowski S. (1996): Neural Networks - An Algorithmic Approach. - Warsaw, Poland: WNT. 
  32. Parisini T., Sanguineti M. and Zoppoli R. (1998): Nonlinear stabilization by receding-horizon neural regulators. - Int. J. Contr., Vol.70, No.3, pp.341-362. Zbl0925.93823
  33. Piche S., Sayyar-Rodsari B., Johnson D. and Gerules M.(2000): Nonlinear model predictive control using neural networks. - IEEE Contr. Syst. Mag., Vol.20, No.3, pp.56-62. 
  34. Qin S. J. and Badgwell T. (2003): A survey of industrial model predictive control technology. - Contr. Eng. Pract.,Vol.11, No.7, pp.733-764. 
  35. Rossiter J. A. (2003): Model-Based Predictive Control. - Boca Raton, FL: CRC Press. 
  36. Sriniwas G. R. and Arkun Y.(1997): A global solution to the non-linear model predictive control algorithms using polynomial ARX models. - Comput. Chem. Engi., Vol.21, No.4, pp.431-439. 
  37. Tatjewski P. (2007): Advanced Control of Industrial Processes, Structures and Algorithms. - London: Springer. Zbl1134.93037
  38. Tatjewski P. and Ławryńczuk M. (2006): Soft computing in model-based predictive control. - Int.J. Appl. Math. Comput. Sci., Vol.16, No.1, pp.101-120. Zbl1334.93068
  39. Trajanoski Z. and Wach P. (1998): Neural predictive control for insulin delivery using the subcutaneous route. - IEEE Trans. Biomed. Eng., Vol.45, No.9, pp.1122-1134. 
  40. Wang L. X. and Wan F. (2001): Structured neural networks for constrained model predictive control. - Automatica, Vol.37, No.8, pp.1235-1243. Zbl0984.93044
  41. Yu D. L. and Gomm J. B. (2003): Implementation of neural network predictive control to a multivariable chemical reactor. - Contr. Eng. Pract., Vol.11, No.11, pp.1315-1323. 
  42. Zheng A. (1997): A computationally efficient nonlinear MPC algorithm. - Proc. American Control Conf., Albuquerque, pp.1623-1627 

NotesEmbed ?

top

You must be logged in to post comments.

To embed these notes on your page include the following JavaScript code on your page where you want the notes to appear.

Only the controls for the widget will be shown in your chosen language. Notes will be shown in their authored language.

Tells the widget how many notes to show per page. You can cycle through additional notes using the next and previous controls.

    
                

                
Note: Best practice suggests putting the JavaScript code just before the closing </body> tag.