A Comparative Analysis of the Performance of Classical Multidimensional Control Systems and Systems with a Neural Controller on the Example of a Real Object
DOI:
https://doi.org/10.24160/1993-6982-2024-2-117-125Keywords:
automatic control systems, neural networks, multidimensionality, automation equipmentAbstract
The purpose of this work is to perform a comparative analysis of two control systems: a control system based on a classical approach in different versions and a control system with a neural controller to evaluate the prospects of using neural networks as a controller.
For performing a comparative analysis, a simulation model of an automatic control system with a multidimensional neural controller is developed with taking into consideration the operation of such automation equipment as a pulse-width modulator and an actuator. The Kashira thermal power plant’s 300 MW power unit No. 3 equipped with a once-through boiler was taken as a control object. Using the optimization algorithm, which is the author's unique development, the neural controller parameters and structure have been found, with which the minimum value of the quadratic integral quality indicator is achieved.
Perturbations were applied to the simulation model via the setpoint adjustment and disturbance channels, control system transients were plotted, and the quality indicators for each perturbation case were calculated.
The newly obtained results were compared with those obtained previously for various structures of control systems with classical controllers. The results are presented in both graphic and tabular forms. The comparative analysis is carried out based on the numerical values of integral quality indicators. A conclusion has been drawn based on the analysis results that neural networks are a promising tool for using as a controller of multidimensional control objects in power engineering.
References
2. Казьмирук И.Ю., Сабанин В.Р., Дементьев Д.А. Применение нейронных сетей в качестве регулятора в двухконтурных системах автоматического регулирования // Молодой ученый. 2019. № 4. С. 91—96.
3. Dementev D.A., Maximova E.D., Sysoletin I.A., Mezin S.V. Research of the Influence of Feedback Signals in the Neuroregulator on the Quality of Regulation // J. Phys.: Conf. Series. 2021. V. 1889(2). P. 022053.
4. Shtylenko A.S., Dementev D.A., Krasnenko D.M., Pomogaev A.S. Exploring the Possibility of Applying Different Neuronal Activation Functions to Single-circuit ACS // Ibid. P. 022007.
5. Shtylenko A.S., Dementev D.A., Krasnenko D.M., Pomogaev A.S. Changing the Pulse Duration of a PWM as a Means of Improving the Quality of the ACS Regulation Using ANN // Ibid. P. 022055.
6. Дементьев Д.А., Максимова Е.Д., Сабанин В.Р., Мезин С.В. Исследование применения искусственного интеллекта в задачах управления многомерными системами // Радиоэлектроника, электротехника и энергетика: Тезисы докл. XXVI Междунар. науч.-техн. конф. М., 2020. С. 924.
7. Жигунов В.В., Шавочкин И.А. Повышение эффективности управления мощностью энергоблоков с применением дифференцирования в котельном регуляторе давления пара перед турбиной // Теплоэнергетика. 2011. № 10. С. 859—868.
8. Свид-во о регистрации программы для ЭВМ № 2021612966. Программа параметрической оптимизации нейрорегулятора для многомерных технологических объектов управления / Дементьев Д.А., Мезин С.В., Максимова Е.Д., Помогаев А.С.
9. Свид-во о регистрации программы для ЭВМ № 2021612965. Программа адаптивной подстройки синаптических коэффициентов нейрорегулятора при динамическом изменении свойств технологического объекта управления / Дементьев Д.А., Мезин С.В., Максимова Е.Д., Помогаев А.С.
10. Биленко В.А., Жигунов В.В. Сравнительный анализ динамических свойств вариантов систем автоматического регулирования частоты и мощности паросиловых энергоблоков с прямоточными котлами // Теплоэнергетика. 2018. № 4. С. 5—20.
---
Для цитирования: Мезин С.В., Дементьев Д.А. Сравнительный анализ работы классических многомерных систем управления и систем с нейрорегулятором на примере реального объекта // Вестник МЭИ. 2024. № 2. С. 117—125. DOI: 10.24160/1993-6982-2024-2-117-125
#
1. Sabanin V.R., Dement'ev D.A., Kaz'miruk I.Yu. Nekotorye Rezul'taty Ispol'zovaniya Iskusstvennykh Neyronnykh Setey v Kachestve Regulyatora Avtomaticheskoy Sistemy Regulirovaniya. Sovremennye Tekhnologii v Zadachakh Upravleniya, Avtomatiki i Obrabotki Informatsii: Sb. Trudov XXVI Mezhdunar. Nauch.-tekhn. Konf. Alushta, 2017:60—61. (in Russian).
2. Kaz'miruk I.Yu., Sabanin V.R., Dement'ev D.A. Primenenie Neyronnykh Setey v Kachestve Regulyatora v Dvukhkonturnykh Sistemakh Avtomaticheskogo Regulirovaniya. Molodoy Uchenyy. 2019;4:91—96. (in Russian).
3. Dementev D.A., Maximova E.D., Sysoletin I.A., Mezin S.V. Research of the Influence of Feedback Signals in the Neuroregulator on the Quality of Regulation. J. Phys.: Conf. Series. 2021;1889(2):022053.
4. Shtylenko A.S., Dementev D.A., Krasnenko D.M., Pomogaev A.S. Exploring the Possibility of Applying Different Neuronal Activation Functions to Single-circuit ACS. Ibid:022007.
5. Shtylenko A.S., Dementev D.A., Krasnenko D.M., Pomogaev A.S. Changing the Pulse Duration of a PWM as a Means of Improving the Quality of the ACS Regulation Using ANN. Ibid:022055.
6. Dement'ev D.A., Maksimova E.D., Sabanin V.R., Mezin S.V. Issledovanie Primeneniya Iskusstvennogo Intellekta v Zadachakh Upravleniya Mnogomernymi Sistemami. Radioelektronika, Eelektrotekhnika i Energetika: Tezisy Dokl. XXVI Mezhdunar. Nauch.-tekhn. Konf. M., 2020:924. (in Russian).
7. Zhigunov V.V., Shavochkin I.A. Povyshenie Effektivnosti Upravleniya Moshchnost'yu Energoblokov s Primeneniem Differentsirovaniya v Kotel'nom Regulyatore Davleniya Para Pered Turbinoy. Teploenergetika. 2011;10:859—868. (in Russian).
8. Svid-vo o Registratsii Programmy dlya EVM № 2021612966. Programma Parametricheskoy Optimizatsii Neyroregulyatora dlya Mnogomernykh Tekhnologicheskikh Ob'ektov Upravleniya. Dement'ev D.A., Mezin S.V., Maksimova E.D., Pomogaev A.S. (in Russian).
9. Svid-vo o Registratsii Programmy dlya EVM № 2021612965. Programma Adaptivnoy Podstroyki Sinapticheskikh Koeffitsientov Neyroregulyatora pri Dinamicheskom Izmenenii Svoystv Tekhnologicheskogo Ob'ekta Upravleniya. Dement'ev D.A., Mezin S.V., Maksimova E.D., Pomogaev A.S. (in Russian).
10. Bilenko V.A., Zhigunov V.V. Sravnitel'nyy Analiz Dinamicheskikh Svoystv Variantov Sistem Avtomaticheskogo Regulirovaniya Chastoty i Moshchnosti Parosilovykh Energoblokov s Pryamotochnymi Kotlami. Teploenergetika. 2018;4:5—20. (in Russian)
---
For citation: Mezin S.V., Dement’ev D.A. A Comparative Analysis of the Performance of Classical Multidimensional Control Systems and Systems with a Neural Controller on the Example of a Real Object. Bulletin of MPEI. 2024;2:117—125. (in Russian). DOI: 10.24160/1993-6982-2024-2-117-125
