SmartGridComm 2018 offers the five tutorials described below. Please notice that attendance requires an additional Half day tutorial registration.
Introduction to Reliable Control using Unreliable Communication Media
This tutorial illustrates topics necessary to know for the realization of reliable control with unreliable communications, aiming to give audience the perspectives to understand the problems and solutions. For industry, it gives a good starting point to solve their problems. Audience from academia will find many interesting technical issues that are still not solved or even not understood well.
In order to realize the reliable remote control, to improve communication quality is important. For this, we need to evaluate communication quality for control applications. Thus, after brief introduction of several important applications of remote control, the tutorial clarifies the differences of conventional and control communications from the viewpoint of performance measures of communication.
It is note worthy that the improvement of communication performance is not the sole solution that is always possible. As an alternative solution, the tutorial introduces the importance and effectiveness of the cross-layer optimization of control and communication, which is important and effective in order to realize reliable control with unreliable communication.
In addition, the talk mentions some interesting research topics on reliable control and communication as open questions for those who are seeking new topics in the field of SG control and communications.
Biography of the speaker:
Masaaki Katayama (SM-IEEE and Fellow-IEICE) was born in Kyoto, Japan in 1959. He received the B.S., M.S. and Ph.D. degrees from Osaka University, Japan in 1981, 1983, and 1986, respectively, all in communication engineering. He was an assistant professor at Toyohashi University of Technology from 1986 to 1989, and a Lecturer at Osaka University from 1989 to 1992. In 1992, he joined Nagoya University as an associate professor, and has been a professor since July 2001.
Prof. Katayama has started his researches on the paticular topics of reliable wireless communications from 1990’s. He is one of the founders of the technical committee on reliable robust radio control (T.C. RRRC) in the Institute of Electronics, Information and Communication (IEICE), Japan, and the previous chair of the committee, which is renamed Technical Committee on Reliable Communication and Control (TC-RCC). He was also a tutorial speaker at Smart Grid Comm. 2011 on PLC with Prof. Andrea Tonello.
General and powerful flexibility modeling and management with flex-offers
As intermittent renewables make up an ever-larger share of the energy production, there is an increasing need to manage demand-side flexible loads in a scalable and effective way. Traditionally, this has mainly been done through implicit models of flexibility, e.g., response to price signals. However, modeling flexibility explicitly enables more of the flexibility to be extracted and utilized.
Many industrial and residential electrical appliances (heat-pumps, batteries, electric vehicles, heating/cooling systems) can be operated flexibly to facilitate grid stabilization, demand-supply balancing, and increased renewable energy integration. Unfortunately, in order to take advantage of such flexible appliances in practice, a new ICT infrastructure, load/optimization models, and accompanying IT tools traditionally needed to be developed for each specific type of appliance (i.e., one for heat-pumps, one for EVs, etc.). In other words, there is a need for a common way of representing and managing energy flexibility.
The ground-breaking flex-offer concept, and its associated techniques and ICT infrastructure, overcomes this problem by providing a common “lingua franca” for flexibility capable of modeling and managing all types of flexible loads, both consumption, production, and storage.
This tutorial will give a detailed description of flex-offers, based on two practical use-cases of loads with continuous (e.g., heat pump) and discrete (e.g., washing machine) state variables. The tutorial will also describe the value of flexibility from these loads for the Nordic day-ahead market (NordPool Spot).
For the two use-cases, the tutorial will focus on two common flexible loads (heat-pumps, washing machines) and walk through the practical steps of modelling, aggregating, optimizing, and disaggregating (the schedules of) these loads modelled as flex-offers. Specifically, the tutorial will show how load-specific “grey-box” models can be created and calibrated using a number of process/user constraints, context information, and power measurements, by utilizing well-known optimization/modelling tools and underlying relevant techniques (e.g., parameter estimation, logistics regression). It will then show how energy flexibility bounds can be estimated using these models and encoded using the generic flex-offer representation. In this representation, thousands of (simple) flex-offers can be efficiently aggregated into one or more (aggregated) flex-offer(-s) with wider flexibility bounds and much higher energy amounts. Thus, an effective representation of an aggregated (super) load is built and efficient optimization of thousands of loads made possible. The tutorial will demonstrate how this aggregation is done and how all the flexible loads in the aggregated form can be efficiently optimized for a given optimization objective using well-known optimization tools and solvers. The outcome of this step is an effective aggregated schedule (a time series), which then will be disaggregated and prescribed to individual flexible loads. It will then demonstrate the disaggregation, in which the aggregated schedule is disaggregated (partitioned) into a number of schedules for each individual load satisfying all load model constraints, ready to be executed by the controllers of these loads. As the last step of this process, this tutorial will demonstrate a real world implementation of the complete flex-offer system, that includes generation, aggregation, scheduling, trading, disaggregation, and execution of flex-offers from real loads.
Biography of the speakers:
Torben Bach Pedersen is Full Professor of computer science at Aalborg University, doing research on big data analytics. He has 300+ peer-reviewed publications with 5600+ citations on Google Scholar (h-index 41). PC Chair and PC member for leading data management conferences. Associate Editor Information Systems and IEEE Transactions on Big Data. ACM Distinguished Scientist and Member of the Danish Academy of Technical Sciences. He (co)led the MIRABEL EU FP7 project which pioneered the flex-offer concept, the large Danish ForskEL Totalflex project which further developed the flex-offer concept and efficient (dis)aggregation techniques, and the Horizon 2020 GOFLEX project which demonstrates flex-offers as the basis for a bottom-up local energy system centered on DSos. He is main supervisor for 12 PhD students and have graduated another 15. His work on flex-offers received the Best Poster Award at World Smart Grid Forum 2013 and Best Paper Award at the ACM International Conference on Future Energy Systems (e-Energy) 2017. He co-leads Center for Data-intensive Systems (40+ researchers), the largest Big Data research group in Denmark. He is co-founder and Chief Scientific Officer of FlexShape, a startup company commercializing the flex-offer technology.
Laurynas Šikšnys is a Postdoc at Aalborg University, doing research on energy data and flexibility management. He received his Bachelor degree in Informatics from Kaunas University of Technology, Lithuania, in 2007. He received an MSc degree in Computer Engineering in 2009; and a Ph.D. degree in Computer Science from Aalborg University and Technische Universität Dresden in 2014. He (co-) developed solutions for flexibility management and trading in the EU FP7 MIRABEL, EU ARTEMIS ARROWHEAD, EU H2020 GOFLEX, and Danish ForskEL TOTALFLEX projects. He received the Best Paper Award at the ACM International Conference on Future Energy Systems (e-Energy) 2017. His additional research interests include real-time data management architectures, optimization problem solving, blockchains, prescriptive analytics, and cyber-physical systems. He is co-founder and CEO of FlexShape, a startup company commercializing the flex-offer technology.
Bijay Neupane is a Postdoc at Aalborg University, doing research on energy analytics. He received an MSc degree in Computer Science from Masdar Institute, UAE, in 2013; and a Ph.D. degree in Computer Science from Aalborg University and Technische Universität Dresden in 2017. His current work focus is on designing and developing solutions for flexibility management and trading. His research interests include predictive modeling, data mining, machine learning, Big data technologies. He has published a number of papers in these fields. He received the Best Paper Award at the ACM International Conference on Future Energy Systems (e-Energy) 2017.
Theoretical foundations for designing an autonomous power grid: PMU data science for blackout and cyber attack early warning
Anticipating power grid instability, voltage collapse, malware attacks, or other threats by Change Point Detection of Phasor Measurement Unit (PMU) signals that identifies the departures from baseline scaling statistics before catastrophic consequences would occur, yet subject to an acceptable False Alarm Rate (FAR), is fundamental for designing an autonomous smart grid. Developing such a data science and statistical signal processing approach would provide the advantage of obviating the need to capture and dissect a worm before countermeasures by signature matching can be deployed. Towards this end, it is critical to mathematically characterize the PMU dynamics, consistent with the fundamental observation that the embedded signals have long-range dependency. This calls for a re-examination of the classical load models and the quest for a paradigm shift in modeling the power grid that could explain the mathematical characteristics of PMU data. Our preliminary studies of the PMU data collected before the Indian grid voltage collapse shows a shift in the frequency data, invisible by naked eye observation of the traces, but upon scaling exponent analysis showing a pronounced increase of the Hurst exponent before the actual collapse occurs. This post-mortem analysis motivates us to develop a real-time Change Point Detection, an algorithm that is very simple to implement, yet relying on advanced statistical concepts such as the CUmulative SUM (CUSUM) or Shriyaev statistic. The proposed methodology computes a CUSUM measuring the departure of the Hurst exponent from its baseline distribution, which if it crosses a threshold it raises an alarm. The major difficulty is to determine such threshold that is consistent with an acceptable FAR in this nonclassical, non Gauss probabilistic context.
Biography of the speakers:
Edmond Jonckheere received the electrical engineering degree from the University of Louvain, Leuven, Belgium; the Dr.-Eng. degree in aerospace engineering from Universit Paul Sabatier, Toulouse, France; and the Ph.D. degree in electrical engineering from the University of Southern California, Los Angeles, CA, USA, in 1973, 1975, and 1978, respectively.
From 1973 to 1975, he was a Research Fellow of the European Space Agency. From 1975 to 1978, he was a Teaching Assistant, Research Assistant, and a Research Associate with the Department of Electrical Engineering— Systems, University of Southern California. In 1979, he was with the Philips Research Laboratory, Brussels, Belgium. In 1980, he was with the University of Southern California, where he is currently a Full Professor of Electrical Engineering and Mathematics, a Member of the Center for Applied Mathematical Sciences, and a Member of the Center for Quantum Information Science and Technology. He held short-term visiting appointments at the Max-Planck-Institute, Gttingen, Germany; Australian National University, Canberra, ACT, Australia; Cardiff University, Wales, U.K.; and Swansea University, Wales. He also held consulting affiliations with the Memorial Medical Center of Long Beach, Long Beach, CA; Lockheed- Martin, Bethesda, MD, USA; the Aerospace Corporation, El Segundo, CA; and Honeywell, Morristown, NJ, USA. His current research interests include conventional versus quantum networks, adiabatic quantum computations, and power grid.
Dr. Jonckheere is a fellow of the Institute of Electrical and Electronics Engineers for contribution to the spectral theory of linear-quadratic and H-Infinity problems.
Paul Bogdan received his Ph.D. degree in electrical and computer engineering from Carnegie Mellon University, Pittsburgh, PA, USA, in 2011. He is currently an Assistant Professor with the Department of Electrical Engineering, University of Southern California, Los Angeles, CA, USA. His current research interests include performance analysis and design methodologies for multicore systems, cyber-physical systems, and modeling and analysis of biological systems.
Laith Shalalfeh received his Ph.D. degree in electrical engineering from University of Southern California, USA. He is currently an Assistant Professor of Electrical Engineering at German Jordanian University.
Convex relaxations and combinatorial optimization of alternating current (AC) electric power systems
Abstract: In the era of dynamic smart grid with fluctuating demands and uncertain renewable energy, it is crucial to continuously optimize the operational cost and performance of electric power grid, while maintaining its state within the stable operating limits. Nonetheless, a major part of electric power grid is consisted of alternating current (AC) electric power systems, which exhibit complex behavior with non-linear operating constraints. The optimization of AC electric power systems is still a very challenging problem that baffles electrical power engineers.
The hardness of optimization problems of AC electric power systems stems from two issues: (1) non-convexity involving complex-valued entities of electric power systems, and (2) combinatoric constraints involving binary control decision variables. Without proper theoretical tools, heuristic methods or general numerical solvers had been utilized traditionally to tackle these problems, which do not provide theoretical guarantees to the true optimal solutions. There have been recent advances in applying convex relaxations to tackle non-convex problems of AC electric power systems. On the other hand, combinatorial optimization is rooted in theoretical computer science, which typically considers linear constraints, instead of those non-linear constraints in AC electric power systems.
This tutorial presents the latest developments in convex relaxation techniques for AC optimal power flow optimization problems with continuous power demands, and the recent extensions to combinatorial optimization of AC optimal power flows with discrete power demands.
Biography of the speakers:
Sid Chi-Kin Chau (https://cecs.anu.edu.au/people/sid-chau) recently joined the Research School of Computer Science at the Australian National University. He was an Associate Professor with the Department of Computer Science at Masdar Institute, which was created in collaboration with MIT, and is a part of Khalifa University of Science and Technology.
His research interests are related to the computing systems and applications for sustainable smart cities, applying Internet-of-things, computational intelligence, advanced algorithms and big data analytics to develop sustainable solutions for smart cities, including smart grid, smart buildings, intelligent vehicles and transportation. He also researches in broad areas of algorithms, optimization, and Internet-of-things.
Previously, he was a visiting professor with MIT, a senior research fellow with A*STAR in Singapore, a Croucher Foundation research fellow with University College London with a research fellowship awarded by the Croucher Foundation Hong Kong, a visiting researcher with IBM Watson Research Center and BBN Technologies, and a post-doctoral research associate with University of Cambridge. He received the Ph.D. from University of Cambridge with a scholarship by the Croucher Foundation Hong Kong, and B.Eng. (First-class Honours) from the Chinese University of Hong Kong.
He is on the TPC of several top conferences in smart energy systems and smart cities, such as ACM e-Energy, ACM BuildSys. He is a TPC co-chair of ACM e-Energy 2018, and guest editor for IEEE Transactions on Sustainable Computing Special Issue on Intersection of Computing and Communication Technologies with Energy Systems and IEEE Journal of IoT Special Issue on Internet-of-Things for Smart Energy Systems. He was a co-founder of a start-up specializing in intelligent systems and big data analytics for smart building management.
Majid Khonji is an assistant professor at Masdar Institute in the Department of Electrical Engineering and Computer Science. He received his BSc degree in software engineering from the UAE University, UAE, in 2007, his MSc degree in Security, Cryptology and Coding of Information Systems from Ensimag, Grenoble Institute of Technology, France, and his PhD degree in Interdisciplinary Engineering from Masdar Institute of Science and Technology in 2016. Previously, he worked as an information security researcher at the Emirates Advanced Investment Group (EAIG), and as a researcher with the Dubai Electricity and Water Authority (DEWA). His research interests are in algorithms, optimization, artificial intelligence, and smart grids.
Smart battery pack
Abstract: Electrical Vehicles (EVs) are emerging on the market driven by the improvements in Li-Ion battery technology. This kind of batteries is the most promising for EV applications due to their long life time, their high energy density and their low self-discharging rate. The recent trend in industry and academy is to conceive new battery packs able to increase the driving range and the battery life time. For these purposes, smart battery cells have bene introduced. They are equipped with a cell management unit that is able to control each cell independently and to perform an active balancing algorithm. The latter enables an efficient balancing of the State of Charge (SoC) among the cells by increasing the actual capacity of the battery pack in both charging and discharging time. The sensing of currents and voltages is also performed inside the smart battery cell and the measured values can be transmitted either to the others cells or to the battery pack controller. This technology allows the integration of all the functionalities (SoC estimation, state of health estimation, SoC balancing) of the Battery Management System (BMS) either at cell level or at pack level.
In  the decentralization of the BMS is achieved by interconnecting all the cells through a communication interface and by implementing an active cell balancing in each smart cell. Authors in  instead propose a centralized BMS based on wireless communication. It leads a simpler, lighter, more reliable and cheaper battery pack. Moreover, no isolators are needed and the wire-harness and connectors are avoided. However, the cell balancing issues is not dealt with. A more recent Smart Battery Pack (SBP) solution has been proposed in . It is also based on wireless communication and a cell bypass balancing technique is presented. This SBP enables all the previous benefits of the wireless BMS by also emphasizing others advantages, such as fault tolerant operation, featuring strong scalability and connection to Internet-of-Things (IoT). However, the wireless communication has different challenges in real EV applications mainly due to the noisy environment, the synchronization requirements and the extremely high dense network (hundreds of cells).
Our tutorial is intended to present the existing solutions and the actual and upcoming challenges in this area. System architectures and hardware implementations are going to be provided along with the future trends in smart battery cells and wireless BMS for EV applications.
S. Steinhorst, “Design and Verification Methodologies for Smart Battery Cells,” in 2016 International Symposium on Integrated Circuits (ISIC), Singapore, 2016.
M. Lee, J. Lee, I. Lee, J. Lee and A. Chon, “Wireless Battery Management System,” in 2013 World Electric Vehicle Symposium and Exhibition (EVS27), Barcelona, 2013.
B. Majmunovic, R. Sarda, R. Teodorescu, C. Lascu and M. Ricco, “Highly Efficient Smart Battery Pack for EV Drivetrains,” in 2017 IEEE Vehicle Power and Propulsion Conference (VPPC), Belfort, 2017.
Remus Teodorescu received the Dipl.Ing. degree in electrical engineering from Polytechnical University of Bucharest, Romania in 1989, and PhD. degree in power electronics from University of Galati, Romania, in 1994. In 1998, he joined Aalborg University, Department of Energy Technology, power electronics section where he currently works as a professor. He is a Fellow Member of IEEE. Between 2013 and 2017, he has been a Visiting Professor with Chalmers University. He was the coordinator of Vestas Power Program (2007 - 2013, involving 10 PhD projects in the areas of power electronics, power systems and energy storage. His areas of interests are: MMC, design and control of power converters used in wind power systems, PV and HVDC/FACTS and smart energy storage systems based on Li batteries and multilevel converters.
Mattia Ricco received the master’s degree (cum laude) in electronic engineering from the University of Salerno, Fisciano, Italy, in 2011, and the Ph.D. degrees in electrical and electronic engineering from the University of Cergy-Pontoise, Cergy-Pontoise, France, and in information engineering from the University of Salerno, Salerno, Italy, in 2015. Since September 2015, he has been a Postdoctoral Research Fellow with Aalborg University, Department of Energy Technology, Aalborg, Denmark. His main research interests include: the design and the implementation of digital control in field programmable gate array, identification techniques for power electronics and photovoltaic systems, control of modular multilevel converters and smart energy storage systems based on Li-ion batteries.