AI and Learning Systems Industrial Applications and Future Directions Edited by Konstantinos Kyprianidis and Erik Dahlquist AI and Learning Systems - Industrial Applications and Future Directions Edited by Konstantinos Kyprianidis and Erik Dahlquist Published in London, United Kingdom Supporting open minds since 2005 AI and Learning Systems - Industrial Applications and Future Directions http://dx.doi.org/10.5772/intechopen.85833 Edited by Konstantinos Kyprianidis and Erik Dahlquist Contributors Weiwei Zhao, Tarek Hassan Mohamed, Gaber Salman, Hussein Abubakr Hussein, Mahmoud Hussein, Gaber Shabib, Andreas Roither-Voigt, Valdemar Lipenko, David Zelenay, Sebastian Nigl, Javad Khazaei, Dinh Hoa Nguyen, Moksadur Rahman, Örjan Larsson, Enislay Ramentol, Tomas Olsson, Shaibal Barua, Idelfonso B. R. Nogueira, Antonio Santos Sánchez, Maria. J. Regufe, Ana M. Ribeiro, Erik Dahlquist, Gladys Bonilla-Enriquez, Patricia Cano Olivos, José Luis Martínez Flores, Diana Sánchez Partida, Santiago Omar Caballero Morales, Karim Belmokhtar, Mauricio Higuita Cano, Jan Skvaril, Konstantinos Kyprianidis, Amare Desalegn Fentaye, Valentina Zaccaria, Ioanna Aslanidou © The Editor(s) and the Author(s) 2021 The rights of the editor(s) and the author(s) have been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights to the book as a whole are reserved by INTECHOPEN LIMITED. The book as a whole (compilation) cannot be reproduced, distributed or used for commercial or non-commercial purposes without INTECHOPEN LIMITED’s written permission. Enquiries concerning the use of the book should be directed to INTECHOPEN LIMITED rights and permissions department (permissions@intechopen.com). Violations are liable to prosecution under the governing Copyright Law. Individual chapters of this publication are distributed under the terms of the Creative Commons Attribution - NonCommercial 4.0 International which permits use, distribution and reproduction of the individual chapters for non-commercial purposes, provided the original author(s) and source publication are appropriately acknowledged. More details and guidelines concerning content reuse and adaptation can be found at http://www.intechopen.com/copyright-policy.html. Notice Statements and opinions expressed in the chapters are these of the individual contributors and not necessarily those of the editors or publisher. No responsibility is accepted for the accuracy of information contained in the published chapters. The publisher assumes no responsibility for any damage or injury to persons or property arising out of the use of any materials, instructions, methods or ideas contained in the book. First published in London, United Kingdom, 2021 by IntechOpen IntechOpen is the global imprint of INTECHOPEN LIMITED, registered in England and Wales, registration number: 11086078, 5 Princes Gate Court, London, SW7 2QJ, United Kingdom Printed in Croatia British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Additional hard and PDF copies can be obtained from orders@intechopen.com AI and Learning Systems - Industrial Applications and Future Directions Edited by Konstantinos Kyprianidis and Erik Dahlquist p. cm. Print ISBN 978-1-78985-877-8 Online ISBN 978-1-78985-878-5 eBook (PDF) ISBN 978-1-83968-601-6 An electronic version of this book is freely available, thanks to the support of libraries working with Knowledge Unlatched. KU is a collaborative initiative designed to make high quality books Open Access for the public good. More information about the initiative and links to the Open Access version can be found at www.knowledgeunlatched.org Selection of our books indexed in the Book Citation Index in Web of Science™ Core Collection (BKCI) Interested in publishing with us? Contact book.department@intechopen.com Numbers displayed above are based on latest data collected. For more information visit www.intechopen.com 5,200+ Open access books available 156 Countries delivered to 12.2% Contributors from top 500 universities Our authors are among the Top 1% most cited scientists 128,000+ International authors and editors 150M+ Downloads We are IntechOpen, the world’s leading publisher of Open Access books Built by scientists, for scientists BOOK CITATION INDEX C L A R I V A T E A N A L Y T I C S I N D E X E D Meet the editors Prof. Konstantinos G. Kyprianidis is a Full Professor in Energy Engineering within the Future Energy Center at Mälardalen University in Sweden. He leads the SOFIA research group (Simu- lation and Optimization for Future Industrial Applications) and is the Head of Research Education for Energy & Environmental Engineering. He has been the Principal Investigator of a large number of national and international research projects relat- ed to automation in the energy and process industry. Among others, he has been the Chief Engineer for the 5.75mEuro project FUDIPO funded by the European Commission. Prior to coming to MDH, he worked for Rolls-Royce plc in the United Kingdom. He has co-authored over 140 peer-reviewed publications and current- ly supervises 15 doctoral candidates and is the Chair of the ASME/IGTI Aircraft Engine Committee. Prof. Erik Dahlquist is a Senior Professor of Energy Technolo- gy and was the former Research Director of the Future Energy Center from 2000 to 2017 at Mälardalen University in Sweden. Among many other projects, he is the Project Coordinator Engi- neer for the 5.75mEuro project FUDIPO funded by the European Commission. Prior to joining MDH, he worked for 28 years at ABB Sweden in various senior technical and management roles. He has co-authored over 300 peer-reviewed publications and currently supervises 10 doctoral candidates. Contents Preface X II I Acknowledgment X V Section 1 Digital Platforms and Learning Systems 1 Chapter 1 3 AI Overview: Methods and Structures by Erik Dahlquist, Moksadur Rahman, Jan Skvaril and Konstantinos Kyprianidis Chapter 2 25 A Framework for Learning System for Complex Industrial Processes by Moksadur Rahman, Amare Desalegn Fentaye, Valentina Zaccaria, Ioanna Aslanidou, Erik Dahlquist and Konstantinos Kyprianidis Chapter 3 55 AI & Digital Platforms: The Market [Part 1] by Örjan Larsson Chapter 4 81 AI & Digital Platforms: The Technology [Part 2] by Örjan Larsson Chapter 5 105 Artificial Intelligence and ISO 26000 (Guidance on Social Responsibility) by Weiwei Zhao Chapter 6 117 Operationalizing Heterogeneous Data-Driven Process Models for Various Industrial Sectors through Microservice-Oriented Cloud-Based Architecture by Valdemar Lipenko, Sebastian Nigl, Andreas Roither-Voigt and Zelenay David Section 2 Industrial Applications of AI 137 Chapter 7 139 Machine Learning Models for Industrial Applications by Enislay Ramentol, Tomas Olsson and Shaibal Barua II Chapter 8 159 Consensus Control of Distributed Battery Energy Storage Devices in Smart Grids by Javad Khazaei and Dinh Hoa Nguyen Chapter 9 175 Power Flow Management Algorithm for a Remote Microgrid Based on Artificial Intelligence Techniques by Karim Belmokhtar and Mauricio Higuita Cano Chapter 10 201 Adaptive Load Frequency Control in Power Systems Using Optimization Techniques by Tarek Hassan Mohamed, Hussein Abubakr, Mahmoud M. Hussein and Gaber S. Salman Chapter 11 217 Modeling the Hidden Risk of Polyethylene Contaminants within the Supply Chain by Gladys Bonilla-Enríquez, Patricia Cano-Olivos, José-Luis Martínez-Flores, Diana Sánchez-Partida and Santiago-Omar Caballero-Morales Chapter 12 231 Sustainable Energy Management of Institutional Buildings through Load Prediction Models: Review and Case Study by Antonio Santos Sánchez, Maria João Regufe, Ana Mafalda Ribeiro and Idelfonso B.R. Nogueira XII Preface Artificial Intelligence (AI) was a “hot” research topic within industrial automation in the early 1980s with developments focused on methods for diagnostics, simula- tion, model adaptation, and optimal control. Artificial neural networks were used extensively at the time within a number of industrial automation applications. Issues included robustness and available computational capacity for online and real-time applications. It took the best part of the next 3 decades to develop practical solutions to these shortcomings, and gradually bring the technology to the level of “intelligence” required to leverage benefits compared to traditional approaches. As a result, interest in the industrial applications of AI and learning systems have surged anew over the last few years. Several powerful methods and digital platforms have been developed during the past decade, and the number of industrial applications has been growing exponentially. This is rapidly changing the way of doing things within the process and energy industry, as well as other industrial sectors. This book covers recent developments and provides a broad perspective of the key challenges that characterize the field of Industry 4.0 with a focus on applications of AI. The target audience for this book includes engineers involved in automation system design, operational planning, and decision support. Computer science practitioners and industrial automation platform developers will also benefit from the timely and accurate information provided in this work. The book is organized into two main sections comprising 12 chapters overall: (i) Digital Platforms and Learning Systems (ii) Industrial Applications of AI The academic editor is indebted to all his colleagues from across the world that contributed to this book with their latest research, to Prof. Erik Dahlquist for joining this effort as co-editor, to several automation experts who volunteered as reviewers, as well as to IntechOpen Publishers for giving me the opportunity to work on this book and its members of staff for their constant support during its preparation. Konstantinos G. Kyprianidis and Erik Dahlquist Professor, Mälardalen University, Västerås, Sweden Acknowledgments This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 723523. Section 1 Digital Platforms and Learning Systems 1 Chapter 1 AI Overview: Methods and Structures Erik Dahlquist, Moksadur Rahman, Jan Skvaril and Konstantinos Kyprianidis Abstract This paper presents an overview of different methods used in what is normally called AI-methods today. The methods have been there for many years, but now have built a platform of methods complementing each other and forming a cluster of tools to be used to build “ learning systems ” . Physical and statistical models are used together and complemented with data cleaning and sorting. Models are then used for many different applications like output prediction, soft sensors, fault detection, diagnostics, decision support, classifications, process optimization, model predictive control, maintenance on demand and production planning. In this chapter we try to give an overview of a number of methods, and how they can be utilized in process industry applications. Keywords: process industry, artificial intelligence (AI), learning system, soft sensors, machine learning 1. Introduction During the 80th AI was a hot topic both in the academia and industries. Many researchers were working a lot with development of methods for diagnostics, sim- ulation and adaptation of models. Artificial Neural Networks (ANN) were being implemented in real applications such as e.g. soft sensors to predict NOx concen- tration in exhaust gas from power plants. Still there was quite some “ over-selling ” and the enthusiasm for AI in the future was assumed to be useful tomorrow. But it took much longer to get the systems robust enough to be used and fast enough to be applicable in on-line applications. After year 2000, systems started to reach a more mature state and we got IBMs Watson, that could beat the Jeopardy master. Later the Google tool could beat the “ Go-master ” , a very complex Chinese game. This has changed the perception of AI. It is still similar type of tools as were developed during the 80th, but now they were refined a lot and hardwires has been developed dramatically. This has given us a much more positive perception of what can be done, and a lot is now being implemented. Still there is a risk for over-selling, as the tools are normally not that “ intelligent ” as we normally think of when we talk about Intelligence. But we are closing the gap day by day. Concerning use of AI in process industry, we cannot just take the tools and hope they will fix everything. It is still important to identify “ what is the problem to solve ” ? With Jeopardy the goal is to be good at Jeopardy, but what is the goal in process industry? It should be to increase production, reduce process variations, 3 implement maintenance on-demand and give operator support. It also means to coordinate and optimize production lines as well as complete plants and later on complete corporations. It also means to adapt to changing customer demands, support in development of new products with production lines as well as handle new business models. These different functions demand quite different tools and thus we will not use only one but several. Often Machine learning is considered being “ the tool ” , but often there is not data available to implement ML, especially not when starting a new production line. To implement new tools, it is also very important to pre-treat data. You have to sort data in “ normal variations ” or “ anom- alies ” . You may need to filter data with moving windows, but in different time perspectives. We need to do data reconciliation to handle drifting sensors. And you need to integrate all levels from orders to production planning down to coordinated and optimized production. In this chapter we will discuss a number of different methods as well as discuss integration between the different levels. Over the years many researchers have investigated different AI techniques for different process industrial application. A comprehensive review on different AI models applied in energy systems can be found in [1]. Applications of different AI tools based on simulation models in pulp and paper industry has been presented by researchers including Dahlquist [2 – 5]. Applications in power plants have been presented in many articles including Karlsson et al. [6 – 8]. In Karlsson et al. [9] a general discussion is made on how to make better use of data including pretreatment of data. Adaptation to degeneration in process models by time is discussed in Karlsson et al. [7]. [10] conducted an extensive review on different AI based soft sensors in process industries. 1.1 Similarities between AI and how the brain works The mathematicians developing especially ANN have been looking a lot on how the brain works. In Figure 1 we see a principal picture of a human. Running in a forest: The brain stores many different factors locally by “ tuning many soft sensors ” . During the night strength of connections are enhanced for the most important functions, while other less important connections are eliminated. Some information is used for direct control. Others is stored for use later on. If it is rainy when you run there is a general feeling that “ this was not so nice ” Everything else happening in the forest then will be “ colored ” by this in your mem- ory, aside of concrete thing like if you meet someone, like a friend, during the run. Figure 1. How a human handle input from the surrounding. 4 AI and Learning Systems - Industrial Applications and Future Directions