=Paper=
{{Paper
|id=Vol-3126/paper56
|storemode=property
|title=Using neural network technologies to simulate the working processes of ship steam boilers
|pdfUrl=https://ceur-ws.org/Vol-3126/paper56.pdf
|volume=Vol-3126
|authors=Vladislav Mikhailenko,Roman Kharchenko,Victor Shcherbinin,Valery Leshchenko
}}
==Using neural network technologies to simulate the working processes of ship steam boilers==
https://ceur-ws.org/Vol-3126/paper56.pdf
Using Neural Network Technologies to Simulate the Working
Processes of Ship Steam Boilers
Vladislav Mikhailenko, Roman Kharchenko, Victor Shcherbinin, Valery Leshchenko.
National University "Odessa Maritime Academy", Didrikhson str.8, Odessa, 65029 Ukraine
Abstract
On the ships of the merchant and passenger fleet, it is relevant to use powerful ship steam boilers
of a wide design class. Marine boilers, as objects of automatic control systems, are subject to
the influence of a significant number of internal and external disturbing factors. Such influences
often lead to self-oscillatory processes of the controlled parameters of a ship's boiler with
significant nonlinearities.
For the optimal tuning of automatic control systems for the working processes of ship boilers,
exact knowledge of mathematical models of controlled processes is required. Due to the
presence of significant nonlinear characteristics, it is proposed to use neural networks in
modeling processes.
As shown by the modeling processes in the MatLab (System Identification Toolbox) program,
the use of nonlinear ARX models with a built-in neural network apparatus makes it possible to
display the experimental working processes of ship parameters with a high degree of adequacy.
Obtaining nonlinear mathematical models with high adequacy will improve the process of
adaptation of automatic control systems for ship boilers and optimize environmental
parameters.
Keywords 1
Steam-boiler, SCADA systems, ARX model, neural network, identification, validation,
neuralnet
1. Introduction tanker "Minerva Roxanne" and obtained using the
monitoring system "ACONIS-2000", are shown
in Fig. 2 [6].
On the ships of the passenger and tanker fleet,
Analysis of the type of transient processes
the technological scheme of operation of two
(Fig. 2) allows us to conclude that with a sharp
auxiliary steam boilers (ASB) and one utilization
increase in the electric and steam load, the turbine
boiler (USB) for a common steam line has found
control system immediately increases steam
wide application (Fig. 1). With such a design
consumption, however, the combustion mode of
solution, auxiliary boilers, performing the
the ASB has not yet been built and an imbalance
function of generating steam of high temperature
occurs in the production and consumption of
and pressure, are subject to the influence of deep
steam, as a result of which the pressure drops.
external disturbances associated with the mode of
steam in the main line and in the path of the
operation of the steam turbine and cargo
working medium of the boiler. An oscillatory
operations on ships [1-5].
mode is formed, characterized by significant
Experimental transient processes of two ASBs
nonlinearity. The ability of the ASB to change the
of Mitsubishi MAC 35 t / h, installed on the oil
ISIT 2021: II International Scientific and Practical Conference
«Intellectual Systems and Information Technologies», September
13–19, 2021, Odesa, Ukraine
EMAIL: vlamihailenod@gmail.com (V. S. Mikhailenko);
romannn30@gmail.com (R. Yu. Kharchenko); lvvlvv@ukr.net
(V. А. Shcherbinin); victor12011201@gmail.com (V. V.
Leshchenko)
ORCID: 0000-0003-2793-8966 (V. S. Mikhailenko); 0000-0003-
3051-7513 (R. Yu. Kharchenko); 0000-0001-6183-5261 (V. А.
Shcherbinin); 0000-0003-0219-5174 (V. V. Leshchenko)
©️ 2021 Copyright for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
steam production in accordance with the change mathematical model for the "fuel consumption -
in the external (electrical) load is called the vapor pressure" channel when the WPC is
maneuverability of the boiler [7]. This condition operating in a maneuverable mode [9].
for the operation of the ABS requires the use of There are two classes of nonlinear regressions
faster-acting ACS so that changes in loads do not [9], with the help of which the nonlinear model of
cause deep deviations in the parameters of the the transient regime is composed (see Fig. 3.10):
working environment. The indicator of the rate of - regressions, non-linear with respect to the
change in the load is the change in pressure in the input and output included in the analysis (explain)
working path of the boiler dР/dt, MPa/min. variables (regressors), but linear in the estimated
parameters (coefficients of the equations);
- regressions, non-linear in the estimated
parameters. For example, the linear structures of
ARX and ARMAX models discussed above can
be extended to nonlinear structures as follows:
- using non-linear ARX regressors, that is,
non-linear expressions of time-delayed input and
output variables;
- replacing the weighted sum of linear
regressors with a nonlinear ARX model, which
has a more flexible nonlinear display function:
F(y(t −1), y(t − 2), y(t −3), …, u(t),
u(t −1),u(t − 2), …),
the arguments forF are the y and u regressor
Figure 1: Control scheme for a group of marine models. For clarity, the nonlinear model of the
auxiliary steam boilers MITSUBISHI MAC 35, ARX structure can be displayed in the block
connected by a common steam pipelines and diagram in Fig. 3.
working on the SHINKO turbine: PC - pressure Using the System Identification Toolbox
regulator; H - remote control; RI - pressure gauge application (Fig. 3 - 4), a discrete ARX [4] model
was obtained, which uses the Z-transform
apparatus:
Discrete-time IDPOLY model:
A(z)y(t) = B(z)u(t) + e(t)
A(z) = 1 + z^-1 + 0.5 z^-2; B(z) = 0.415;
е(t) –discrete white noise, where z-1 = e-sT is the
delay operator; T - sampling interval.
Figure 2: Transient characteristics of the WPC
Figure 3: Block diagram of a nonlinear ARX
during parallel operation of the oil tanker
model
"Minerva Roxanne" at the SPTU: N - load; Рп -
steam pressure; RT - fuel pressure [8]
The process of determining the degree of
2. Development of mathematical adequacy of the selected model is shown in Fig.
models 5. According to the analysis of the degree of
adequacy of the analyzed models, calculated in
Nonlinear models are used to compile a
the software application (see Fig. 5), it was found operate on the common steam line of the turbine,
that the ARX model demonstrates the highest using the SCADA monitoring system ACONIS-
degree of convergence with the experimental 2000E, installed in the central control room of the
ship, the experimental characteristics obtained
data.
(Fig. 6 - 7).
For the process under study - two auxiliary boilers
installed on the tanker "Minerva Roxanne", which
Figure 4: Nonlinear ARX model describing the process of changing the vapor pressure of the ASB
Figure 5: The investigated process in the System Identification Toolbox application obtained on the
basis of the used nonlinear model
Figure 6: Transient characteristics of ACS content of O2 in the exhaust gases of the combined (1) and
auxiliary (2) ASB of the oil tanker "Minerva Roxanne": T - time constant, Kfix - coefficient, z - delay
Figure 7: Screenshot from the mnemonic diagram showing the change in the feed water flow rate in
the combined (1) and auxiliary (2) ASB, working together in transient modes on the tanker "Minerva
Roxanne"
It should be noted that control objects 3. Review of the validation process
demonstrate significant nonlinear characteristics,
therefore, to obtain a model of the system under
consideration, a nonlinear ARX model was used. The process of identification and validation
on an independent data set in the System
Identification Toolbox (SIT) is shown in Fig. 8 -
9.
Figure 8: Experimental dependence of the oxygen content in the exhaust gases
Figure 9: Determining the degree of adequacy of nonlinear models in the System Identification
Toolbox during verification: nlarx1 - nonlinear ARX model with a degree of adequacy of 88.8%
In fig. 10-11 show the view of the nonlinear determined using the SIT application.
model and its three-dimensional surface, as
Figure 10: Structure and parameters of non-linear ARX model
Figure 11: Three-dimensional surface of the acquired non-linear ARX model
It should be noted that the System network (neuralnet), and a linear estimation
Identification Toolbox provides several (linear) [10-11]. By default, a non-linear
nonlinear estimates of g (x) for nonlinear ARX estimation in the form of a wavelet is used (see
models. Nonlinearity is formed in the form of a Fig. 10).
wavelet, a sigmoid network (sigmoidnet), a
binary tree (treepartition), a multilevel neural
4. Conclusions [7] Marine Boiler and Steam Turbine Generator.
URL:
https://www.mhimme.com/auxiliary_boilers
Based on the study, the results were obtained
.html.
that make it possible to improve the toolkit for
[8] Mikhailenko, V. S., & Kharchenko, R. Y.
using the nonlinear ARX model in the form of a
(2014). Analysis of traditional and neuro-
multilevel neural network and a linear assessment
fuzzy adaptive system of controlling the
for parametric identification of the oxygen
primary steam temperature in the direct flow
content characteristic in the exhaust gases from
steam generators in thermal power stations.
the thermal load of the ASB, which makes it
Automatic Control and Computer Sciences,
possible to display the process under study with a
48(6),334–344.
degree of adequacy equal to 95%, and use the
doi:10.3103/s0146411614060066.
obtained model of a high degree of adequacy for
[9] Mitsubishi Auxiliary Boiler MAC-B. URL:
analyzing the process of the appearance of oxygen
https://ru.scribd.com/document/334763233/
corrosion in the equipment of a ship's boiler.
Mitsubishi-Auxiliary-Boiler-MAC-B-pdf.
[10] Mikhaylenko, V. S., Kharchenko, R. Y., &
5. References Shcherbinin, V. A. (2020). Analysis of the
Predicting Neural Network Person
[1] Wei Wang, Han-Xiong Li, Jingtao Zhang. Recognition System by Picture Image.
Intelligence-Based Hybrid Control for Power Automatic Control and Computer Sciences,
Plant Boiler // IEEE Transactions on Control 54(3),249–258.
Systems Technology, Vol. 10, No. 2 March doi:10.3103/s0146411620030037.
2012. – P. 10 –18 [11] Pham, D. T., & Liu, X. (1995). Neural
[2] Xiangjie Liu, Felipe Lara-Rosano. Networks for Identification, Prediction and
Generalized Minimum Variance Control of Control. doi:10.1007/978-1-4471-3244-8.
Steam-Boiler Temperature using Neuro-
Fuzzy Approach // Proceedings of the
5'WorId Congress on Intelligent Control and
Automation, Hangzhou, P.R. China, June 15
– 19, 2018. – P. 134 – 148.
[3] Xiangjie Liu, Hao Zhang. Modeling of an
ultra-supercritical boiler-turbine system with
stacked denoising auto-encoder and long
short-term memory network // Information
Sciences, Vol. 525, Juil 2020. – P. 134 – 152.
[4] Yu Yanzhi, Xu Haichuan, Lv Junxian, Chen
Duoggang ,Dong Gongjun,Shen Bo, Liu
Sheng. 600 MW Supercritical Boiler’s Main
Steam Temperature Control Analysis //
WASE International Conference on
Information Engineer-ing, 2016. – P. 56 – 68.
[5] Gu Xiu Ping. Study on the Superheated
Steam Temperature Control System Based
on Fuzzy-Neural Network // Tianjin
University, 2017. – P. 118 –126.
[6] Mikhailenko, V. S., Shcherbinin V. A.,
Leshchenko, V. V. Kharchenko, R. Yu., N.
V. Lozhechnikova (2020) Modeling the
process of hazardous emissions formation in
the exhaust gases of ship's steam boilers
Informatics and Mathematical Methods in
Simulation Vol. 10 (2020), No. 3-4, pp.
154-166. doi 10.15276/imms.v10.no 3-
4.154.