O.V. Nepomnyashchiy1, I.A. Rusak2, N.Y. Sirotinina3, A.A. Kopytov4, V.N. Khaidukova5

1–5 Siberian Federal University (Krasnoyarsk, Russia)
 

The article introduces the electric motor intelligent control using the reference motor model based on a neural network (neuroemulator). The advantages of the proposed approach are the versatility of the control system, its adaptation to any motor type, and eliminating the rotor shaft speed sensor of the electric motor. To generate a training sample, the motor model has been developed in the MatLab software environment. The neuroemulator of an electric motor has been implemented using the recurrent ANN NARX. The Levenberg-Marquardt method was used for training. The trained neural network is embedded in the developed model of the electric motor control loop. The results of modeling an intelligent control system showed good compliance of the data generated by the neuroemulator with real life data produced by the electric motor.

Nepomnyashchiy O.V., Rusak I.A., Sirotinina N.Y., Kopytov A.A., Khaidukova V.N. Model of the adaptive system based on an artificial neural network for digital electric motor control. Science Intensive Technologies. 2023. V. 24. № 6. P. 43−51. DOI: https://doi.org/10.18127/ j19998465-202306-05

1. Bobtsov A.A., Pyrkin A.A. Adaptive and Robust Control with Uncertainty Compensation. SPb.: NRU ITMO. 2013. 135 p.
2. Alexandrov A.G., Palenov M.V. State and prospects of development of adaptive PID controllers in technical systems. Automation and telemechanics. 2014. V. 2. P. 16–30.
3. Åström K.J., Hägglund T. Advanced PID control (V. 461). Research Triangle Park, NC: ISA-The Instrumentation, Systems, and Automation Society, 2006. 461 p.
4. Sidorova A.A., Malyshenko A.M. Analysis of the effectiveness of algorithms for automatic tuning of adaptive industrial PID-controllers. Bulletin of the Tomsk Polytechnic University. 2011. V. 318. № 5. P. 5–6.
5. Glushchenko A.I. A method for determining the learning rate of a neural network for the problem of on-line adjustment of linear controllers when controlling nonlinear objects. Stary Oskol: MISIS branch. 2018. 107 p.
6. Terekhov V.M., Osipov O.I. Electric drive control systems. 2nd ed. Moscow: Academy. 2006. 304 p.
7. Udut L.S., Maltseva O.P., Koyain N.V. Design and research of automated electric drives. Asynchronous frequency-controlled electric drive. Tomsk: TPU. 2010. 448 p.
8. Langraf S.V., Glazyrin A.S., Afanasyev K.S. The use of the Luenberger observer for the synthesis of vector sensorless asynchronous electric drives. Proceedings of higher educational institutions. Electromechanics. 2011. № 6. P. 57–61.
9. Langraf S.V., Glazyrin A.S., Glazyrina T.A., Afanasyev K.S., Timoshkin V.V., Kozlova L.E. Investigation of parametric robustness of sensorless vector asynchronous electric drive with Kalman identifier. Bulletin of the Tomsk Polytechnic University. 2010. V. 317. № 4. P. 120–123.
10. Vinogradov A.B. Vector control of alternating current electric drives. Ivanovo: Ivanovo State Power Engineering University n. a. V.I. Lenin. 2015. 298 p.
11. Ismeal G.A., Kyslan K., Fedák V. DC motor identification based on Recurrent Neural Networks. Proceedings of the 16th International Conference on Mechatronics-Mechatronika. 2014. Dec 3. Technical University of Košice, Slovak Republic. IEEE. 2015. P. 701–705.
12. Nesterov K.E. Development and research of the system “Thyristor voltage converter – asynchronous motor” with a calculator of the rotor speed according to the EMF of the stator (Ph.D. Thesis). Ekaterinburg. 2009. 140 p.
13. Kozlova L.E., Bolovin E.V. Study of the statics and dynamics of a closed sensorless asynchronous electric drive for auxiliary needs of TPPs according to the TRN – IM scheme with a neural network observer of angular velocity. Modern problems of science and education. 2014. № 3. P. 6.
14. Boussaada Z., Curea O., Remaci A., Camblong H., Mrabet Bellaaj N. A nonlinear autoregressive exogenous (NARX) neural network model for the prediction of the daily direct solar radiation. Energies. 2018. V. 11. № 3. P. 620.
15. Costa M.A., de Pádua Braga A., de Menezes B.R. Improving generalization of MLPs with sliding mode control and the Levenberg–Marquardt algorithm. Neurocomputing. 2007. V. 70. № 7–9. P. 1342–1347.
16. Gavin, H.P. The Levenberg-Marquardt algorithm for nonlinear least squares curve-fitting problems. Department of Civil and Environmental Engineering, Duke University. 2007. P. 1–19.