Prospects for control methods in engineering systems

Cover Page

Cite item

Full Text

Open Access Open Access
Restricted Access Access granted
Restricted Access Subscription Access

Abstract

This article highlights the prerequisites and natural effects of control method development in engineering systems: (1) a simple deviation and perturbation controller, (2) a fuzzy logic controller with a fuzzifier and rule base, (3) a neural network controller for dynamically adjusting the coefficients of the corresponding links, (4) a discrete neural network controller with a neural approximator and controller. The experience gained by researchers and engineers since the initial description of regulatory principles in 1910, including the level of information technology design, particularly the neural network approach to machine learning and the enormous computing potential of computer devices, now enable the integration of a fundamentally novel method of discrete neural network regulation.

The article’s review aims to identify and demonstrate the importance of experimental and operational data, which must be organized and annotated at the time of collection and archiving. This approach will allow us to rapidly implement neural network controllers in engineering systems, as the most critical phase in their development is involves learning and optimization of neural network architecture.

The article presents the principle of operation, benefits, and drawbacks, and the optimal stages for enhancing a neural network controller based on two neural networks, which form a control strategy while considering the most probable state of the system at the next point in time.

About the authors

Vladislav M. Mamedov

Bauman Moscow State Technical University

Author for correspondence.
Email: mamedov-vm@bk.ru
ORCID iD: 0009-0004-8780-7401
SPIN-code: 4095-0195

Postgraduate Student, Assistant Lecturer

Russian Federation, Moscow

Ivan A. Arkharov

Bauman Moscow State Technical University

Email: arkharov@bmstu.ru
ORCID iD: 0000-0002-1624-171X
SPIN-code: 9674-4585

Dr. Sci. (Tech.), Professor

Russian Federation, Moscow

References

  1. Mamedov V, Arkharov I, Navasardyan E. Concept design of cryogenic system of the SPD-detector for NICA project in Dubna. In: IIR Conference: The 16th Cryogenics 2021, October 5–7, 2021. 2021:24–29. doi: 10.18462/iir.cryo.2021.0005
  2. Denisenko VV. PID-reguljatory: principy postroenija i modifikacii. Sovremennye tehnologii avtomatizacii. 2006. № 4. S. 66–74. (In Russ).
  3. Dormido S. Advanced PID Control, IEEE Control Systems Magazine. 2006;26(1):98–101. doi: 10.1109/MCS.2006.1580160
  4. Ang KH, Chong G, Li Y. PID control system analysis, design, and technology. IEEE Transactions on Control Systems Technology. 2005;13(4):559–576.
  5. Ziegler JG, Nichols NB. Optimum settings for automatic controllers. Trans. ASME. 1942;64:759-768.
  6. Cai J. A Fully Mechanical Realization of PID Controller. Highlights in Science, Engineering and Technology. 2022;9:319–328. doi: 10.54097/hset.v9i.1861
  7. Egupova ND. Metody robastnogo, nejro-nechjotkogo i adaptivnogo upravlenija: uchebnik. Moscow: MGTU im. Baumana; 2002. (In Russ).
  8. Uskov AA, Kuz’min AV. Intellektual’nye tehnologii upravlenija. Iskusstvennye nejronnye seti i nechetkaja logika. Moscow: Gorjachaja linija-Telekom; 2004. (In Russ).
  9. Demidova GL, Lukichev DV. Reguljatory na osnove nechetkoj logiki v sistemah upravlenija tehnicheskimi obtektami. Saint Petersburg: Universitet ITMO; 2017. (In Russ).
  10. Pérez-Gomariz M, López-Gómez A, Cerdán-Cartagena F. Artificial Neural Networks as Artificial Intelligence Technique for Energy Saving in Refrigeration Systems - A Review. Clean. Technol. 2023;5:116–136. doi: 10.3390/cleantechnol5010007
  11. Wilson CL, Wilkinson RA and Garris MD, Self-organizing neural network character recognition on a massively parallel computer. In: IJCNN International Joint Conference on Neural Networks, San Diego, CA, USA. 1990;2:325–329. doi: 10.1109/IJCNN.1990.137734
  12. Shalf J. The future of computing beyond Moore’s Law. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences. 2020;378(2166). doi: 10.1098/rsta.2019.0061
  13. Zubkova VV. Analiz aktual’nosti zakona Mura. Perspektivy razvitija informacionnyh tehnologij. 2014;21:136–1140. (In Russ).
  14. Guzhva A, Dolenko S, Persiantsev I. Multifold Acceleration of Neural Network Computations Using GPU. In: Alippi C, Polycarpou M, Panayiotou C, Ellinas G. (eds) Artificial Neural Networks – ICANN 2009. ICANN 2009. Lecture Notes in Computer Science. Berlin, Heidelberg: Springer; 2009;5768. doi: 10.1007/978-3-642-04274-4_39
  15. Benderskaja EN, Tolstov AA. Tendencii razvitija sredstv apparatnoj podderzhki nejrovychislenij. Nauchno-tehnicheskie vedomosti SPbGPU. Informatika. Telekommunikacii. Upravlenie. 2013;3(174):9–18. (In Russ).
  16. Kyoung-Su Oh KS, Jung K. GPU implementation of neural networks. Pattern Recognition. 2004;37:1311–1314.
  17. Bouzar-Benlabiod L, Rubin SH and Benaida A. Optimizing Deep Neural Network Architectures: an overview. In: IEEE 22nd International Conference on Information Reuse and Integration for Data Science (IRI), Las Vegas, NV, USA. 2021:25–32. doi: 10.1109/IRI51335.2021.00010
  18. Ünal HT, Başçiftçi F. Evolutionary design of neural network architectures: a review of three decades of research. Artif. Intell. Rev. 2022;55:1723–1802. doi: 10.1007/s10462-021-10049-5
  19. Dzhulli A, Pal S. Biblioteka Keras – instrument glubokogo obuchenija. Realizacija nejronnyh setej s pomoshh’ju bibliotek Theano i TensorFlow. Moscow: DMK Press; 2018. (In Russ).
  20. Golovko VA. Nejrosetevye tehnologii obrabotki dannyh: ucheb. Posobie. Minsk: BGU; 2017. (In Russ).

Supplementary files

Supplementary Files
Action
1. JATS XML
Download
2. Fig. 1. Structural diagrams of regulators based on the principles of deviation (а) and perturbation (b). Designations: R – controller, OR – control object, D – sensor, Xs – target parameter setting, XOR – control target parameter, XD – sensor readings, ex – control error, F – external influence, and G – control action.

Download (85KB)
Indexing metadata
3. Fig. 2. Structural diagram of a controller with fuzzy logic. Designations on the diagram: OR – object of regulation, D – sensor, BP – rule base, >Ф – fuzzifier, Ф> – defuzzifier, L – logical conclusion, Xs – target parameter setting, XOR – target control parameter, XD – sensor readings, ex – regulation error, Ф – regulation phase, F – external action, G – control action.

Download (53KB)
Indexing metadata
4. Fig. 3. Structural diagram of a PID controller with a neural network for adjusting the coefficients. Designations in the diagram: R – controller, OR – control object, D – sensor, NW – neural network, Xs – target parameter setting, XOR – target control parameter, XD – sensor readings, ex – control error, F – external influence, G – control action, K, Ti, Td – the corresponding coefficients of the PID controller.

Download (49KB)
Indexing metadata
5. Fig. 4. Block diagram of a discrete neural network controller. Designations: OR – control object, D – sensor, NWA – neural network approximator, NWC – neural network controller, Xs – target parameter setting, XOR – control target parameter, ex – control error, S – control strategy by control action, С – executive device, F – external action, G – control action.

Download (126KB)
Indexing metadata

Copyright (c) 2023 Mamedov V.M., Arkharov I.A.

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Согласие на обработку персональных данных

 

Используя сайт https://journals.rcsi.science, я (далее – «Пользователь» или «Субъект персональных данных») даю согласие на обработку персональных данных на этом сайте (текст Согласия) и на обработку персональных данных с помощью сервиса «Яндекс.Метрика» (текст Согласия).