Radiotekhnika
Publishing house Radiotekhnika

"Publishing house Radiotekhnika":
scientific and technical literature.
Books and journals of publishing houses: IPRZHR, RS-PRESS, SCIENCE-PRESS


Тел.: +7 (495) 625-9241

 

Electric gas-turbine unit control system automation tuning on the neural network basis

DOI 10.18127/j20700814-201809-05

Keywords:

B.V. Kavalerov - Dr.Sc.(Eng.), Associate Professor, Head of Department «Electrical Engineering and Electrical Mechanics», Perm National Research Polytechnic University
E-mail: kbv@pstu.ru
G.A. Kilin - Master, Senior Lecturer, Department «Electrical Engineering and Electrical Mechanics»,  Perm National Research Polytechnic University
E-mail: thisisforasm@rambler.ru


To use GTU as part of the power plant, an automatic control system (ACS) is needed, which should provide all the main indicators of the quality of electricity generation. The problem is that basically these automatic control systems are transferred practically unchanged from their aviation prototypes. As a consequence, the electric power system (EPS) dynamics is not taken into account at all. This situation leads to a decrease in the efficiency of the decisions made at the design stages of the automatic control system, which in turn leads to serious obstacles in ensuring the necessary performance characteristics of gas turbine power plants (GTPPs). The main reason for the situation arises in the absence of adequate software and algorithmic tools that provide the solution of the tasks of complex automation of the debugging and tuning of ACS for ground-based electric power GTUs.
The technological process of tuning the ACS GTU is an integral part of the GTES tests. Particularly costly are the operations of manual adjustment of the automatic control system for full-scale tests, since in this case the tuning time substantially increases with simultaneous increase in the cost of operation of the GTES. In addition, the ACS GTU itself is synthesized without taking into account the dynamics of the power system, which also has a negative impact on the process of electricity generation. That is why it is advisable to preliminarily use the mathematical model of GTU and EPS to pre-configure the automatic control system, which makes it possible to promptly set up or test a new ACS, while spending minimum efforts and time. As a result, the process of setting up the ACS consists in the following: first, the procedure for setting up the ACS on the mathematical model is performed, then the received settings of the automatic control system are tested on a semi-stand, finally, the decisions taken at the field stand or under the direct operation of the GTES are checked.
Unfortunately, most existing mathematical models are not suitable for automating the adjustment of the parameters of the controllers of the automatic control system of the GTU, since they are either designed for a different purpose, or their speed is too low. That is why it is necessary to develop a neural network mathematical model for automating the adjustment of the parameters of the GTU regulator and on its basis to propose a new technique for tuning the ACS by electric power GTUs. Such a developed model should include not only the GTU, but also the dynamics of EPS, it should be simplified, which is necessary to achieve an acceptable simulation speed and automatic tuning. Such a model should have sufficient accuracy to obtain on this model the settings of the controllers of the automatic control system.
On work the following conclusions are made:
1) On the need for a model-oriented approach for the automatic control systems synthesis at the stage of scientific research trials.
2) On the possibility of using a neural network approach for constructing gas turbine power plant model.

References:
  1. Inozemczev A.A., Nixamkin M.A. i dr. Avtomatika i regulirovanie aviaczionny’x dvigatelej i e’nergeticheskix ustanovok. T. 5. M.: Mashinostroenie. 2008. 190 s.
  2. Adgamov R.I., Berxeev M.M., Zalyaev I.A. Avtomatizirovanny’e ispy’taniya v aviastroenii. M.: Mashinostroenie. 1989. 232 s.
  3. Rotach V.Ya., Kuzishhin V.F., Klyuev A.S. i dr. Avtomatizacziya nastrojki sistem upravleniya / Pod red. V.Ya. Rotacha. M.: E’nergoatomizdat. 1984. 272 s.
  4. Shigapov A.A., Petrochenkov A.B., Kavalerov B.V. Strukturnaya sxema matematicheskoj modeli sistemy’ e’lektrosnabzheniya dlya ispy’taniya algoritmov upravleniya // E’lektrotexnika. 2010. № 6. S. 55−61.
  5. Venikov V.A. Teoriya podobiya i modelirovanie: Ucheb. posobie dlya vuzov. Izd. 2-e, dop. i pererab. M.: Vy’sshaya shkola. 1976. 479 s.
  6. Kavalerov B.V., Kazanczev V.P. Matematicheskaya model’ e’lektroe’nergeticheskoj sistemy’ dlya optimal’noj nastrojki regulyatorov gazoturbinny’x mini-e’lektrostanczij // Vestnik Izhevskogo gosudarstvennogo texnicheskogo universiteta. 2011. № 2(50). S. 21−24.
  7. Kavalerov B.V., Petrochenkov A.B. Modelirovanie e’lektricheskoj nagruzki dlya nastrojki sistem upravleniya konvertirovanny’mi gazoturbinny’mi ustanovkami // E’lektrotexnika. 2011. № 11. S. 11−16.
  8. Kilin G.A., Kavalerov B.V. Perspektivy’ ispol’zovaniya nejrosetevy’x texnologij v zadache polucheniya matematicheskix modelej sistemy’ «gazoturbinnaya ustanovka – sinxronny’j generator» // Materialy’ II Mezhdunar. nauchno-texnich. konf. «Avtomatizacziya v e’lektroe’nergetike i e’lektrotexnike». 21−22 aprelya 2016 g. Perm’: Izd-vo PNIPU. S. 51−55.
  9. Zhdanovskij E.O., Kavalerov B.V., Kilin G.A. Razrabotka nejrosetevoj modeli gazoturbinnoj e’lektrostanczii dlya nastrojki regulyatorov gazoturbinnoj ustanovki // Fundamental’ny’e issledovaniya. 2016. № 12-3. S. 479−485.
  10. Kilin G.A., Kavalerov B.V., Masyagin E.D. Vy’bor arxitektury’ nejronnoj seti dlya postroeniya matematicheskoj modeli gazoturbinnoj e’lektrostanczii // Materialy’ konf. «Aktual’ny’e problemy’ e’lektromexaniki i e’lektrotexnologij». UrFU. Ekaterinburg. 2017. S. 209−214.
June 24, 2020
May 29, 2020

© Издательство «РАДИОТЕХНИКА», 2004-2017            Тел.: (495) 625-9241                   Designed by [SWAP]Studio