Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
4.5
2.7
Journals
Applied Sciences
Special Issues
Modeling and Simulation with Artificial Neural Network
share announcement

Modeling and Simulation with Artificial Neural Network

A special issue of Applied Sciences (ISSN 2076-3417). This special issue belongs to the section "Computing and Artificial Intelligence".

Deadline for manuscript submissions: closed (31 December 2021) | Viewed by 4851

Special Issue Editors

Prof. Dr. Tag Gon Kim

E-Mail Website
Guest Editor
School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon 34141, Korea
Interests: modeling and simulation theory/applications; machine learning embedded modeling and simulation; digital twins
Prof. Seon Han Choi

E-Mail Website
Guest Editor
Department of IT Convergence and Application Engineering, Pukyong National University, Busan 48513, Korea
Interests: discrete event system simulation; simulation-based optimization; machine learning; digital twins

Special Issue Information

Dear Colleagues,

Modeling and Simulation (M&S) is a powerful way to study and analyze complex systems. From a classical M&S perspective, a system model stands for an abstract expression of the operational principles or knowledge of the system with differential equations, tables, algorithms, or formal specifications (e.g., DEVS and Petri Net). Recently, as deep learning has been emerged, a data-driven approach to modeling such systems with an input–output relationship via artificial neural network (ANN) is drawing attention. Compared to the classical modeling, this approach has an advantage that can develop a high-fidelity model from the collected big data of a system without the explicit knowledge of the system. However, the trained ANN has limitations in prediction accuracy when an operational environment of the system is changed such as a change of an input domain, system’s structure and/or component values.

Since the classical M&S and the data-driven approach are complementary, using both of them together can develop a more flexible and reliable model for the complex system compared to using only one. For example, when developing a model for optimal control of a smart greenhouse, two components should be considered: 1) a controller and 2) the greenhouse. The classical M&S can be a better approach to developing a model of the controller because the controller has a clear operational principle. On the other hand, an accurate model for the green house which represents the dynamic behavior for changes of environmental parameters, such temperature and humidity, is not clear. Fortunately, such a model can be obtained as an ANN data model using big data accumulated over a considerable time. As a result, a flexible and high-fidelity model for the smart greenhouse can be developed by combining these two models.

This Special Issue covers the overall research fields related to a complementary use of M&S and ANN, ranging from concepts, theories, methodologies, and applications to practical studies in specific domains. Given your renowned expertise and significant contributions to this field, we would like to invite you to contribute to this Special Issue.

Prof. Dr. Tag Gon Kim
Prof. Seon Han Choi
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Applied Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • Modeling and Simulation
  • Artificial Neural Network
  • Data-driven model
  • Big Data

Published Papers (3 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

24 pages, 3530 KiB  
Article
Advanced Machine Learning Applications to Viscous Oil-Water Multi-Phase Flow
by Sayeed Rushd, Uneb Gazder, Hisham Jahangir Qureshi and Md Arifuzzaman
Appl. Sci. 2022, 12(10), 4871; https://0-doi-org.brum.beds.ac.uk/10.3390/app12104871 - 11 May 2022
Cited by 10 | Viewed by 1328
Abstract
The importance of heavy oil in the world oil market has increased over the past twenty years as light oil reserves have declined steadily. The high viscosity of this kind of unconventional oil results in high energy consumption for its transportation, which significantly [...] Read more.
The importance of heavy oil in the world oil market has increased over the past twenty years as light oil reserves have declined steadily. The high viscosity of this kind of unconventional oil results in high energy consumption for its transportation, which significantly increases production costs. A cost-effective solution for the long-distance transport of viscous crudes could be water-lubricated flow technology. A water ring separates the viscous oil-core from the pipe wall in such a pipeline. The main challenge in using this kind of lubricated system is the need for a model that can provide reliable predictions of friction losses. An artificial neural network (ANN) was used in this study to model pressure losses based on 225 data sets from independent sources. The seven input variables used in the current ANN model are pipe diameter, average velocity, oil density, oil viscosity, water density, water viscosity, and water content. The ANN developed using the backpropagation technique with seven processing neurons or nodes in the hidden layer demonstrated to be the optimal architecture. A comparison of ANN with other artificial intelligence and parametric techniques shows the promising precision of the current model. After the model was validated, a sensitivity analysis determined the relative order of significance of the input parameters. Some of the input parameters had linear effects, while other parameters had polynomial effects of varying degrees on the friction losses. Full article
(This article belongs to the Special Issue Modeling and Simulation with Artificial Neural Network)
►▼ Show Figures

Figure 1

17 pages, 2638 KiB  
Article
Comparing Polynomials and Neural Network to Modelling Injection Dosages in Modern CI Engines
by Tomasz Osipowicz, Karol Franciszek Abramek and Łukasz Mozga
Appl. Sci. 2022, 12(4), 2246; https://0-doi-org.brum.beds.ac.uk/10.3390/app12042246 - 21 Feb 2022
Cited by 1 | Viewed by 1282
Abstract
The article discusses the possibility of using computational methods for modelling the size of the injection doses. Polynomial and artificial intelligence methods were used for prediction. The aim of the research was to analyze whether it is possible to model the operating parameters [...] Read more.
The article discusses the possibility of using computational methods for modelling the size of the injection doses. Polynomial and artificial intelligence methods were used for prediction. The aim of the research was to analyze whether it is possible to model the operating parameters of the fuel injector without knowing its internal dimensions and tribological associations. The black box method was used to make the model. This method is based on the analysis of input and output parameters and their correlation. The paper proposes a mathematical model determined on the basis of a polynomial and a neural network based on input and output parameters. The above models make it possible to predict the amount of fuel injection doses on the basis of their operating parameters. Modelling was performed in the Matlab environment. Calculating methods could support the diagnosis processes of fuel injectors. Fuel injection characteristic is non-linear. Study shows that it is possible to predict injection characteristic with high matching using polynomial and neural network. That way accelerates fuel injector work parameters research process. Fuel injector test basis on known its work areas. Mathematical modelling can calculate all injection area using few parameters. To modelling fuel injection dosages by neural network have been used back propagation and Levenberg—Marquardt algorithms. Full article
(This article belongs to the Special Issue Modeling and Simulation with Artificial Neural Network)
►▼ Show Figures

Figure 1

25 pages, 3579 KiB  
Article
A Convex Combination Approach for Artificial Neural Network of Interval Data
by Woraphon Yamaka, Rungrapee Phadkantha and Paravee Maneejuk
Appl. Sci. 2021, 11(9), 3997; https://0-doi-org.brum.beds.ac.uk/10.3390/app11093997 - 28 Apr 2021
Cited by 3 | Viewed by 1567
Abstract
As the conventional models for time series forecasting often use single-valued data (e.g., closing daily price data or the end of the day data), a large amount of information during the day is neglected. Traditionally, the fixed reference points from intervals, such as [...] Read more.
As the conventional models for time series forecasting often use single-valued data (e.g., closing daily price data or the end of the day data), a large amount of information during the day is neglected. Traditionally, the fixed reference points from intervals, such as midpoints, ranges, and lower and upper bounds, are generally considered to build the models. However, as different datasets provide different information in intervals and may exhibit nonlinear behavior, conventional models cannot be effectively implemented and may not be guaranteed to provide accurate results. To address these problems, we propose the artificial neural network with convex combination (ANN-CC) model for interval-valued data. The convex combination method provides a flexible way to explore the best reference points from both input and output variables. These reference points were then used to build the nonlinear ANN model. Both simulation and real application studies are conducted to evaluate the accuracy of the proposed forecasting ANN-CC model. Our model was also compared with traditional linear regression forecasting (information-theoretic method, parametrized approach center and range) and conventional ANN models for interval-valued data prediction (regularized ANN-LU and ANN-Center). The simulation results show that the proposed ANN-CC model is a suitable alternative to interval-valued data forecasting because it provides the lowest forecasting error in both linear and nonlinear relationships between the input and output data. Furthermore, empirical results on two datasets also confirmed that the proposed ANN-CC model outperformed the conventional models. Full article
(This article belongs to the Special Issue Modeling and Simulation with Artificial Neural Network)
►▼ Show Figures

Graphical abstract

Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-3
Back to TopTop