Article

Article title ON DEVELOPMENT OF METHOD TO CALCULATE TIME DELAY VALUES OF NEURAL NETWORK INPUT SIGNALS TO IMPLEMENT PI-CONTROLLER NEURAL TUNER
Authors Y.I. Eremenko, D.A. Poleshchenko, A.I. Glushchenko
Section SECTION IV. NEURAL NETWORK CONTROL ALGORITHMS
Month, Year 10, 2015 @en
Index UDC 004.89 + 681.51
DOI
Abstract A neural tuner is used to increase energy efficiency of unsymmetrical plants described by first or second order aperiodic links with time delay. It allows to tune KP and KI parameters of a PI-controller online without knowledge of plants model. A main part of the tuner is a neural network, which input vector includes plant output value signals delayed on equal time gaps Δt from each other. The main aim of the research is to develop a method to calculate Δt depended on plant parameters values. More than 15000 experiments are conducted with plant models using different values of time constant, plant gain and delay time. Δt values are changed from 1 s to 40 s for each certain model. The best experiment is chosen from the set of 40 on the basis of proposed criteria. Such experiment shows the best value of Δt for the plant model in question. Having conducted experiments, sought analytical dependence is found. It is also shown that the number of the neural tuner calls N during each transient of each experiment with the best value of Δt is a constant. The neural tuner is called every Δt seconds. It is also shown that N value probability curve is Gaussian distribution. Such facts are proved theoretically. On that base a method to calculate Δt value without knowledge about plant parameters values is proposed. Further research needs to be done to include parameter Δt into neural network operative training process to be able to refine it during control system functioning. In that case obtained analytical dependence will be used to initialize Δt.

Download PDF

Keywords Adaptive control; neural networks; PI-controller; neural tuner; input signals time delay.
References 1. Astrom K.J., Hagglund T., Hang C.C., Ho W.K. Automatic tuning and adaptation for PID controllers. A survey, IFAC J. Control Eng. Practice, 1993. – Vol. 1, No. 4. – P. 699-714.
2. Astrom K.J., Hagglund T. Advanced PID Control. Research Triangle Park: ISA – The Instrumentation, Systems, and Automation Society, 2006, 461 p.
3. Alexandrov A.G., Palenov M.V Self–tuning PID–I controller: Proceedings of the 18th IFAC World Congress. Milano, Italy, 2011, pp. 3635-3640.
4. Pfeiffer B.-M. Towards «plug and control»: self–tuning temperature controller for PLC, International journal of Adaptive Control and Signal Processing, 2000, No. 14, pp. 519-532.
5. Ziegler J., Nichols N. Optimum settings for automatic controllers, Trans. ASME, 1942, No. 65, pp. 759-768.
6. Shubladze A.M., Kuznetsov S.I. Avtomaticheski nastraivayushchiesya promyshlennye PI i PID regulyatory [Automatically tunable industrial PI and PID controllers], Avtomatizatsiya v promyshlennosti [Automation in Industry], 2007, No. 2, pp. 15-17.
7. Rotach V.Ya. Kuzishchin V.F., Petrov S.V. Nastroyka regulyatorov po perekhodnym kharakteristikam sistem upravleniya bez ikh approksimatsii analiticheskimi vyrazheniyami [Controllers adjustment based on control system transient characteristics without their approximation with analytical expressions], Avtomatizatsiya v promyshlennosti [Automation in Industry], No. 11, 2009, pp. 9-12.
8. Chen J., Huang T. Applying neural networks to on-line updated PID controllers for nonlinear process control, Journal of Process Control, 2004, No. 14, pp. 211-230.
9. Fang M., Zhuo Y., Lee Z. The application of the self–tuning neural network PID controller on the ship roll reduction in random waves, Ocean Engineering, 2010, No. 37, pp. 529-538.
10. Anderson K.L., Blankenship G.I., Lebow L.G. A rule–based adaptive PID controller, Proc. 27th IEEE Conf. De-cision. Control, 1988, pp. 564-569.
11. Omatu S., Khalid M., Yusof R. Neuro-Control and its Applications. London: Springer, 1995, 255 p.
12. Reyes J., Astorga C., Adam M., Guerrero G. Bounded neuro–control position regulation for a geared DC motor, Engineering Applications of Artificial Intelligence, 2010, No. 23, pp. 1398-1407.
13. Tan S.–H., Hang C.–C., Chai J.–S. Gain scheduling: from conventional to neuro–fuzzy, Automatica, 1997, Vol. 33, No. 3, pp. 411-419.
14. Zhao Z.Y., Tomizuka M., Isaka S. Fuzzy gain scheduling of PID controllers, IEEE Transactions on systems. man. and cybernetics, 1993, No. 5, pp. 1392–1398.
15. Unal M., Ak A., Topuz V., Erdal H. Optimization of PID Controllers Using Ant Colony and Genetic Algorithms. London: Springer, 2013, 85 p.
16. Eremenko Yu.I., Poleshchenko D.A., Glushchenko A.I. O primenenii neyrosetevogo optimizatora parametrov PI-regulyatora dlya upravleniya nagrevatel'nymi pechami v razlichnykh rezhimakh raboty [About usage of PI-controller parameters neural tuner for control of heating furnaces functioning in different modes], Upravlenie bol'shimi sistemami [Large-Scale Systems Control], 2015, Issue 56, pp.143-175.
17. Hagglund T., Tengvall A. An automatic tuning procedure for unsymmetrical processes, Proceedings of 3rd European control conference, 1995, pp. 2450-2455.
18. Eremenko Yu.I., Poleshchenko D.A., Glushchenko A.I., Fomin A.V. Ob otsenke primenimosti razlichnykh struktur neyronnoy seti v realizatsii neyrosetevogo optimizatora parametrov PI-regulyatora dlya upravleniya teplovymi ob"ektami [On applicability estimation of different neural networks structures for implementation of PI-controller parameters neural optimizer for heating plants control], Sistemy upravleniya i informatsionnye tekhnologii [Management Systems and Information Technology], 2014, No. 3.2 (57), pp. 236-241.
19. Hornik K., Stinchcombe M., White H. Multilayer feed-forward networks are universal approximators, Neural networks, 1989, No. 2 (5), pp. 359-366.
20. Huang G.B., Wang D.H., Lan Y. Extreme learning machines: a survey, International Journal of Machine Learning Cybernetics, 2011, No. 2, pp. 107-122.

Comments are closed.