Rahman Khorramfar - EIMC2 Seminars

This page contains the past and upcoming seminars organized by the Energy & Infrastructure Systems: Modeling, Computing, and Control (EIMC2) group at the Laboratory for Information and Decision Systems (LIDS), MIT.

These seminars provide a platform for researchers and practitioners whose work broadly fits within the scope of EIMC2 to disseminate their ideas, receive feedback, and connect with colleagues. The presentation format is in-person, but a Zoom link is available to join virtually. We record presentations with the speakers' permission and archive them at 'EIMC2 Seminars' channel on YouTube.

Please reach out to the seminar co-organizers, Rahman Khorramfar (khorram@mit.edu) or Luis Mansilla (luiscvm@mit.edu), for questions, comments, or speaker suggestions.

Upcoming Seminar(s)

(May 8) Conleigh Byers: Electricity Market Design and Risk Trading with Heterogeneous Demand

Date and Location: May 8, 2025, 11am at MIT 45-600B

Speaker: Conleigh Byers, Harvard University

Abstract: The rise of variable renewable energy, storage, and load growth due to data centers with increasing extreme weather events has called into question whether current electricity market designs are sufficient for a decarbonized future. Many systems today have turned to some form of centrally administered resource adequacy mechanism beyond short-run markets to manage risk. Energy consumers are heterogeneous – differentiated by demand profiles, flexibility, reliability requirements, and risk appetite. The primary contributions of this work are (1) formulation of a non-algorithmic method for incomplete markets with risk-trading (2) incorporation of a multitude of swap and option contract designs (3) incorporation of endogenous demand in market participation and risk trading. From this framework, we can explore which contracts designs are appropriate for which types of heterogeneous demand. These conclusions should in turn influence the ongoing debate on organized long-run markets for electricity.

Speaker Bio: Conleigh Byers is an Environmental Fellow at the Harvard University Center for the Environment and the Harvard Kennedy School of Government. Her research uses tools from operations research, electrical engineering, and economics to design decarbonized energy systems, with a focus on electricity market design and power systems operations and planning. She received a doctorate in Electrical Engineering from ETH Zürich, a dual masters from MIT in Electrical Engineering & Computer Science and Technology & Policy, and an undergraduate degree from Princeton University.

(May 15) Pablo Ducru: Techno-economics of the energy transition -- a macro perspective

Date and Location: May 15, 2025, 11am at MIT 45-600B

Speaker: Pablo Ducru, MIT

Abstract: This lecture seeks to bring clarity to the climate change problem. We cover the key factors at play – from physics, to technology, to economics – showing how population, growth, the economy-energy coupling, and the carbon-intensity of the energy system, contribute to greenhouse gas emissions and rising temperatures. This will make clear the levers we have – from zero-carbon energy, to efficiency gains, to carbon pricing – so you can focus your attention on the most impactful solutions. This will clarify how you can contribute to solving climate change: modeling and monitoring the climate; deploying existing technologies (solar, wind, hydro, nuclear, to grid and energy infrastructures — supply, demand, transmission, efficiency, and storage), building the regulatory frameworks and leveraging finance and markets to create economic incentives with system-wide impact (and for different income-level countries), or developing the new technologies needed (energy, materials, industry, adaptation, etc.).

Speaker Bio: Pablo Ducru is a Franco-Mexican scientist, and entrepreneur. Pablo is a lecturer at MIT. Prior to this, Pablo co-founded Raive, an artificial intelligence (AI) company building vision-language foundation models to power the world’s content, enabling large-scale training, hosting, and servicing of custom models, so that everyone can have and exchange their own personalized AI models – an infrastructure upon which a market of AIs with IP attribution and AI royalties can emerge. Pablo is a world-expert in nuclear data and developing algorithms for high-performance computations. He received his PhD in computational nuclear science and engineering, worked at Los Alamos and Oak Ridge National Labs, developing new nuclear physics and statistical learning models for high-performance diffusion models. Pablo is also a Schwarzman Scholar from Tsinghua University. He conducted quantitative finance research for energy markets, and was advisor to the Minister of Economy of Mexico. In France, Pablo graduated from École Polytechnique (X-2011), where he served as an Officer in the Navy.

Past Seminars (reverse chronological order)

13- Lingling Fan: Connect Control Theory with the Practical World: Replication of power grid dynamic events and control design for inverter-based resources (Slides)

Date and Location: May 1, 2025, 11am at MIT 45-600B

Speaker: Lingling Fan, University of South Florida

Abstract: In this talk, the speaker will share her experiences on connecting control theory for real-world problem solving, including replication of power grid dynamic events and control design for inverter-based resources. While control theory is well-developed, implementing it for real-world problem-solving requires integrating it with domain knowledge. Creating abstractions, formulating linear time-invariant (LTI) systems, and designing feedback systems all require domain knowledge. The first example examines subsynchronous resonance events that led to rotor shaft damage in synchronous generators. The second example explores control design for inverters to achieve key control functions while ensuring tight current regulation. In both cases, LTI systems are constructed for modeling and analysis. Numerical demonstrations will highlight the effectiveness of classic control theory.

Speaker Bio: Dr. Lingling Fan is Professor at the Department of Electrical Engineering at the University of South Florida. Prior to her academic career, she was with Midwest ISO as a transmission planning engineer. Dr. Fan is research active in control, system identification, and stability analysis of power systems, power electronic converters, and electric machines. Dr. Fan is the founding co-chair of IEEE Power and Energy Society’s Wind SSO/IBR SSO task force. She has authored/co-authored three books on dynamic modeling of power grids, synchronous machines, and inverter-based resources. Dr. Fan served as the Editor-in-Chief of IEEE Electrification Magazine from 2020 to 2024 and Consulting Editor of IEEE transactions on Sustainable Energy from 2018 to 2020. Currently, she serves in the steering committee of IEEE Electrification Magazine and the editorial board of IEEE transactions on Energy Conversion. Dr. Fan was elevated to IEEE Fellow class 2022 for her contributions to stability analysis and control of inverter-based resources. She received Energy Systems Integration Group (ESIG) Excellent Award for her contributions to power system dynamics and system oscillations in IBR penetrated power grids in March 2024.

12- Mehdi Jafari: Grid Planning for Reliability: Rethinking ELCC

Date and Location: April 10, 2025, 11am at MIT 45-600B

Speaker: Mehdi Jafari, Staff Software Engineer, Form Energy

Abstract: Modern electricity systems demand both cost-effectiveness and high reliability, especially with increasing renewable energy integration. Systems planners must navigate complex conditions such as decarbonization goals, variable weather patterns, and increasing demands to ensure continuous reliable supply. A key metric for reliability is the Effective Load Carrying Capability (ELCC) which has inherent complexities: it's highly sensitive to interdependent factors, such as system configuration, installed capacity, resource diversity, and ordering effect. Traditional methods use generic ELCC values from a third party that can result in suboptimal investment decisions. To address these limitations, this talk presents a novel multi-scenario co-optimization model. This framework integrates capacity expansion, production cost model, and reliability constraints, enabling the endogenous calculation of capacity accreditations and eliminating the need for predefined ELCC values. By directly co-optimizing economic and reliability subproblems, this approach overcomes dimensionality challenges and the ordering effect, delivering improved accuracy and speed, and ultimately facilitating the development of robust and least-cost electricity portfolios.

Speaker Bio: Mehdi Jafari is a Staff Software Engineer – Modeling & Optimization – at Form Energy with over ten years of research and work experience in power and energy systems. He holds a Ph.D. in Electrical Engineering and was a Postdoctoral Associate at LIDS (2018-2022). Dr. Jafari's work focuses on developing advanced energy storage and renewable energy solutions for modern energy systems, employing optimization algorithms, probabilistic analysis, and techno-economic decision-making. At Form Energy, he is developing high-fidelity energy storage system models, granular electricity system models for reliability and integrated resource planning studies, and models for time-domain reduction and spatiotemporal flexibility. He has authored over 40 publications in peer-reviewed journals and conferences.

11- Michael Davidson: Optimizing Land and Grid Impacts of Renewables Dominant Power Systems (Video)

Date and Location: March 26, 2025, 11am at MIT 45-600B

Speaker: Michael Davidson, UC San Diego, UCSD

Abstract: Widespread use of renewable electricity sources is necessary to address climate change, but their intermittency, geospatial variability, and large land footprints create unique challenges for optimizing deployment of continental systems. In this talk, I will describe the key computational trade-offs and optimization methods to plan large-scale power systems and present a modeling framework to consider spatially granular land and grid impacts. Applying this to China, a plausible scenario leads to 2,000 and 3,900 GW of wind and solar capacity, respectively, to reach carbon neutrality in 2060. Competition for alternative uses of land, notably agriculture, lead to key trade-offs between technology choice and location. Temporal scale-up and institutional limitations create additional barriers to building renewable dominant systems by mid-century. New areas of model development that combine engineering and political economy insights will be discussed.

Speaker Bio: Michael Davidson is an assistant professor joint with the Mechanical and Aerospace Engineering Department and School of Global Policy and Strategy at the University of California, San Diego, where he leads the Power Transformation Lab. Dr. Davidson’s teaching and research focus on the engineering implications and institutional conflicts inherent in deploying low-carbon energy at scale to mitigate environmental harms. Michael holds a Ph.D. in engineering systems and an S.M. in technology and policy from MIT, and a B.S. in mathematics and physics and a B.A. in Japanese studies from Case Western Reserve University. He has held fellowships at the Harvard Kennedy School and Tsinghua University, and worked on U.S.-China climate policy for the Natural Resources Defense Council.

10- Pablo Ruiz: Efficient and Reliable Congestion Mitigation with Topology Optimization: Experience in the Midcontinent ISO Footprint (Video, Slides)

Date and Location: March 13, 2025, 11am at MIT 45-600B

Speaker: Pablo Ruiz, NewGrid, The Brattle Group and Boston University

Abstract: Timely and effective management of transmission system limitations has become critical to support the ongoing energy transition as well as economic development. Additional transmission capability is needed quickly to support increased deployment of variable energy resources, which often are located away from demand centers, the electrification of traditionally fossil energy uses, and other load growth, such as data centers. Transmission congestion can become acute in many regions, and has imposed over $20 billion of added costs on US transmission customers in 2022. It also impacts grid reliability and resilience throughout the country. To cost-effectively manage transmission constraints will require improved congestion management practices and technologies to better utilize the existing grid, especially as new transmission facilities are planned and constructed. Transmission topology optimization identifies grid reconfiguration options to quickly re-route power flow around bottlenecks, increasing transfer capability across congested areas. These reconfiguration solutions reduce the market redispatch and generation curtailments otherwise necessary to manage congestion and, by doing so, provide operational, efficiency, reliability, and resilience benefits. To facilitate the use of reconfigurations and provide another congestion management tool to transmission customers, the Midcontinent ISO (MISO) has implemented a process for market participants to submit reconfiguration proposals that show promise after initial analysis. The Southwest Power Pool (SPP) and the Electric Reliability Council of Texas (ERCOT) have similar process initiatives under discussion. This presentation will demonstrate the reliability, cost-saving, environmental and curtailment-mitigation impacts of reconfigurations identified on behalf of market participants in MISO, as well as illustrate additional benefits that could be obtained if topology optimization opportunities were used proactively and consistently across RTO footprints.

Speaker Bio: Pablo A. Ruiz is Co-founder and CEO of NewGrid, a grid optimization company, where he leads the teams that develop topology optimization software technology and provide congestion monitoring and mitigation services to utilities and market participants. He is also a senior consultant at The Brattle Group and Associate Research Professor of Mechanical Engineering at Boston University. He was the Principal Investigator for the U.S. DOE ARPA-E Transmission Topology Control Algorithms project, which led to the formation of NewGrid. Pablo holds a Ph.D. in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign.

9- Line Roald: Carbon-Sensitive Electric Loads and their Impacts on Emissions (Video, Slides)

Date and Location: March 6, 2025, 11am at MIT 45-600B

Speaker: Line Roald, University of Wisconsin-Madison

Abstract: An increasing number of electric loads can be characterized as "carbon-sensitive", meaning that they are willing to adapt their real-time electricity use in response to the current carbon intensity of the grid. Examples of carbon-sensitive consumers include hyperscale computing companies, who want to publicly demonstrate their commitment to clean energy, and hydrogen producers, who are incentivized by tax credits or valuable "green" classifications. Individual consumers, often on flat electricity tariffs, simply want to contribute to a cleaner planet. While there is a lot of interest in "doing the right thing" to reduce the carbon footprint of our electricity use, there is concerningly little understanding of what the "right" action for end consumers actually is. We will discuss our analysis of common carbon emission metrics and demonstrate their sometimes counterintuitive and counterproductive impacts on grid emissions. We will then discuss some more holistic approaches to allow loads to expose carbon preferences to the grid.

Speaker Bio: Line Roald is an Associate Professor in the Department of Electrical and Computer Engineering at University of Wisconsin—Madison. She received her Ph.D. degree in Electrical Engineering (2016) from ETH Zurich, Switzerland, and was a postdoctoral research fellow at Los Alamos National Laboratory. She is the recipient of an NSF CAREER award, the Vilas Early Career Investigator Award and several best paper awards. Her research interests center around modeling and optimization of energy systems, with a particular focus on managing uncertainty and risk from extreme weather and renewable energy variability.

8- Emiliano Dall'Anese: Constrained Online Feedback Optimization: Recent Advances and Applications to Power Systems (Video)

Date and Location: Feb 27, 2025, 11am at MIT 45-600B

Speaker: Emiliano Dall'Anese, Boston University

Abstract: Online feedback optimization (OFO) is a framework to design feedback controllers that regulate a system – either dynamical or modeled by an algebraic map - toward equilibria defined as optimal solutions of an optimization problem. By leveraging the interplay of OFO and control barrier functions (CBFs), this talk presents new strategies for the design of feedback controllers that drive the system to solutions of optimization problems with steady-state constraints on both inputs and states. In this talk, we first present a new state feedback controller based on CBF tools and a continuous approximation of the projected gradient flow; we show that the equilibria of the interconnection between the plant and the proposed controller correspond to critical points of the constrained (possibly nonconvex) optimization problem. Sufficient conditions to ensure that, for the closed-loop system, isolated locally optimal solutions of the optimization problem are locally exponentially stable are presented, and forward invariance of the set of feasible inputs is established. Then, we present an application of the proposed method in the context of power distribution systems; in this application, the OFO method is utilized to solve an AC optimal power flow (OPF) problem based only on voltage measurements (and without requiring measurements of the power consumption of non-controllable assets). We show that the proposed method leads to significantly less voltage violations in a regime with high renewable generation relative to competing alternatives.

Speaker Bio: Emiliano Dall'Anese is an Associate Professor in the Department of Electrical and Computer Engineering at Boston University, where he is also an appointed faculty with the Division of Systems Engineering. He is an affiliate faculty with the Center for Information & Systems Engineering and the Institute for Global Sustainability. He received the Ph.D. in Information Engineering from the Department of Information Engineering, University of Padova, Italy, in 2011. He was with the University of Minnesota as a postdoc (2011-2014), the National Renewable Energy Laboratory as a senior researcher (2014-2018), and the Department of Electrical, Computer, and Energy Engineering at the University of Colorado Boulder as a faculty (2018-2024). His research interests span the areas of optimization, control, and learning; current applications include power systems and autonomous systems. He received the National Science Foundation CAREER Award in 2020, the IEEE PES Prize Paper Award in 2021, the IEEE Transactions on Control of Network Systems Best Paper Award in 2023, and the IEEE PES ISGT Europe best paper award in 2024.

7- Rahman Khorramfar: Robust Power-Gas Infrastructure Planning under Supply and Demand Uncertainties (Slides)

Date and Location: Nov 26, 2024, 4pm at MIT 45-500A

Speaker: Rahman Khorramfar, MIT

Abstract: Achieving economy-wide decarbonization via variable renewable energy (VRE) expansion and electrification of end-uses requires new approaches for energy infrastructure planning that consider, among other factors, weather-driven uncertainty in demand and VRE supply. A planning model that fails to account for these uncertainties can hinder the intended transition efforts and increase the risk of disruptions in energy supply, especially during extreme weather conditions. Here, we propose a distributional robust optimization (DRO) approach for the generation and transmission expansion problem (GTEP) of joint power-gas infrastructure and operations planning under the uncertainty of both demand and renewable supply. Aiming to construct a more realistic approach while avoiding significant computational burden, we embed the spatial correlation in both electric power and gas networks as part of the ambiguity set design. Furthermore, we consider the risk-awareness of the energy planners in the modeling framework via the conditional value-at-risk (CVaR) metric.

Speaker Bio: Rahman Khorramfar is currently a postdoctoral associate at MIT Energy Initiative and LIDS, developing advanced optimization and machine learning tools for design, planning, and operations of climate-resilient energy systems. He received PhD in Industrial and Systems Engineering from NC State University where he primarily worked on hierarchical decision-making and optimization under uncertainty and their applications in production planning, supply chain, and capacity expansion problems.

6- Deepjyoti Deka: Fast Data-Driven Optimization and Risk Estimation Using Limited Data in Power Grids (Video, Slides)

Date and Location: Nov 12, 2024, 4pm at MIT 45-500A

Speaker: Deep Deka, MIT

Abstract: Fast optimization of generator dispatch and associated risk estimation are crucial to ensuring reliable and economical delivery of electricity in power grids. AC Optimal Power Flow (OPF), with non-linear AC power flow equations and constraints, is the primary optimization framework used for power grid operational and planning problems. In this talk, we present results on two Bayesian data-driven proxies to learn AC power flow equations and AC -OPF surrogates using limited data, with a focus on probabilistic predictions.

The first data-driven model uses graph-structured Gaussian process (GP) model for risk assessment of critical voltage constraints. The proposed GP is based on a novel kernel, named the vertex-degree kernel (VDK), that decomposes the voltage-load relationship based on the network graph, and is estimated using Active Learning. Simulations demonstrate that the proposed VDK-GP achieves a reduction in computational time by 15 times over Monte- Carlo simulations in estimating system risk.

The second model termed BayesOPF is a Bayesian Neural Networks (BNN) based proxy for AC-OPF using limited OPF training time and samples. BayesOPF is trained in a sandwiched manner that alternates between a supervised learning step for minimizing cost, and an unsupervised learning step for enforcing feasibility. We show that BayesOPF outperforms conventional neural network architectures on OPF problems, without requiring any correction or projection step.

Speaker Bio: Deepjyoti Deka is a research scientist at MIT Energy Initiative. His research interests lie at the intersection of machine learning and optimization for tractable sensing and secure operation in renewable rich power grids. From 2018-2024, he was a staff scientist in the Theoretical Division at Los Alamos National Laboratory and served as a PI/co-PI for DOE and internal projects on machine learning in power systems and optimization of interdependent networks. He received the M.S. and Ph.D. degrees in Electrical and Computer Engineering (ECE) from the University of Texas, Austin, TX, USA, in 2011 and 2015, respectively. He completed his undergraduate degree in Electronics and Communication Engineering (ECE) from IIT Guwahati, India with an institute silver medal as the best outgoing student of the department in 2009. Dr. Deka is a senior member of IEEE, and has served as an editor on IEEE Transactions on Smart Grid.

5- Nicolas Stevens: On Some Advantages of Convex Hull Pricing for the European Electricity Auction (Video, Slides)

Date and Location: Oct 29, 2024, 4pm at MIT 45-500A

Speaker: Nicolas Stevens, UCLouvain, Belgium & Harvard

Abstract: Since the liberalization of the power sector and the creation of wholesale electricity markets, the question of how to price the non-convexities that are present in the market has attracted the interest of both academics and practitioners. Over the years, US markets have adopted different and evolving pricing rules which are still vividly debated. Since the “Trilateral Market Coupling” (2006), the European day-ahead market has opted for a notably different pricing rule, and there is currently research undertaken by the EU stakeholders to reform it. Our work aims at contributing to the debate. We analyze six different pricing methods. We establish several mathematical properties for enabling their comparison. Our findings are illustrated on stylized examples and numerical simulations that are performed on realistic auction datasets. Both theoretical and numerical evidences that are gathered in our paper point towards the advantages of the so-called “convex hull pricing” approach.

Speaker Bio: Nicolas Stevens is a PhD candidate at the Center for Operations Research and Econometrics (CORE), UCLouvain, Belgium. He is also currently a visiting researcher at the Harvard Kennedy School. His PhD dissertation, Price Formation with Non-Convexities : Theory and Applications for the Electricity Market, examines how the pricing rules in electricity auctions may accommodate the non-convexities that are present in the market orders. His research interests include energy economics, energy policy, market design, auctions and operations research. He also holds a Master of Engineering in Applied Mathematics from UCLouvain. In the past, Nicolas worked in the industry as a consultant in the energy sector on various topics including the design of the European day-ahead market in collaboration with the European Power Exchanges.

4- Yuanyuan Shi: Learning for Power Grid and Building Control (Video, Slides)

Date and Location: Oct 28, 2024, 11am at MIT 45-500A

Speaker: Yuanyuan Shi, UC San Diego

Abstract: In this talk, I will share our recent progress on developing learning algorithms for real-world energy system control, with stability and computational tractability guarantees.

The first application is power grid voltage control. I will introduce a novel neural network architecture – monotone neural network (MNN) that ensure the network output is a monotone function of the input. MNN is achieved by first designing neural networks that are convex (with universal approximation guarantee) and using gradients of convex functions to ensure monotonicity. We show that MNN is a powerful structure for voltage control – with stability guarantee and superior performance compared to standard neural networks.

The second application is building control. There is an emergent need to model indoor air quality to improve occupant health and building energy efficiency. A fundamental challenge is that building airflow dynamics are governed by nonlinear partial differential equations (PDEs) with unknown parameters, which are computationally prohibitive from a real‑time control perspective. I will introduce our work on PDE‑constrained optimization for building model identification and designing neural operator learning for efficient PDE system control.

Speaker Bio: Yuanyuan Shi is an Assistant Professor at the Department of Electrical and Computer Engineering at the University of California San Diego. She received her Ph.D. in Electrical and Computer Engineering (ECE) from the University of Washington in 2020. From 2020 to 2021, she was a Postdoctoral Scholar at Caltech. Her research focuses on machine learning, dynamical systems and control, with applications to sustainable energy systems. She is a recipient of the Hellman Fellowship in 2023, the UW Clean Energy Institution Scientific Achievement Award in 2020, and best paper finalist from ACM e-Energy 2022.

3- Hannele Holttinen & Julia Matevosyan: Recommended Practices for Wind and Solar Integration Studies (Video, Slides)

Date and Location: Oct 25, 2024, 11am at MIT 45-322

Speakers: Hannele Holttinen (Aalto University) & Julia Matevosyan (Energy Systems Integration Group, ESIG

Abstract: The International Energy Agency (IEA) WIND Task 25 ”Design and Operation of Energy Systems with Large Amounts of Variable Generation” has compiled Recommended Practices for power system impact studies, commonly known as wind and solar integration studies. It provides research institutes, consultants, and system operators with the best available information on how to perform an integration study, considering resource adequacy, operational, and stability challenges. It is also useful in benchmarking any integration studies: the recommendations check list can be used to identify what has and has not been taken into account. The latest update, to Edition 3, includes recommendations for very high wind and solar shares– wind and solar dominated power systems, with sector coupling and energy system integration.

Speaker Bio:

Hannele Holttinen is currently Professor of Practice at Aalto University, Finland, as well as RDI Lead, Energy systems at CLIC Innovation Oy. Through her own consultancy Recognis Oy she continues coordinating international research on grid integration at IEA WIND (Task 25), and at G-PST (Global Power System Transformation Consortium). She worked previously at VTT Technical Research Centre of Finland in different fields of wind energy and energy system integration research.

Julia Matevosyan is a Chief Engineer at Energy Systems Integration Group (ESIG) and has more than 20 years of experience in the power industry. Prior to joining ESIG, Julia was the Lead Planning Engineer of the Electric Reliability Council of Texas (ERCOT). In her time with ERCOT, she worked on adequacy of system inertial response, system flexibility, frequency control and performance issues related to high penetration levels of inverter-based generation and ancillary services market design. Julia received her BSc from Riga Technical University in Latvia, and her MSc and PhD from the Royal Institute of Technology (KTH) in Sweden

2- Mark O’Malley: Inertia in Power Systems and in Power Systems Research (Video)

Date and Location: Oct 18, 2024, 4pm at MIT 45-332

Speaker: Mark O’Malley, Imperial College London

Abstract: Connection to the power system via synchronous machines (SMs) was the norm for fossil fuels, nuclear, hydro etc. and they inherently have inertial response. Variable renewable energy (VRE) sources (wind & solar photovoltaic) are connected to the grid with power electronic (inverters/converters). These inverter-based resources (IBRs) do not behave like SMs and e.g. have no inertial response. With increasing penetrations of VRE inertia in power systems is now a very popular research topic. The impact on power systems of this trend and its impacts on power systems and the urgent need for research will be explored. When coupled with “inertia” in the power systems research community some interesting observations can be made.

Speaker Bio: Mark O’Malley is the Leverhulme Professor of Power Systems at Imperial College London. He is a co-founder of the Global Power System Transformation Consortium, Global Director for the international research centre, Electric Power for a Carbon free Society and a Hagler Fellow at Texas A&M. He is a Foreign Member of the US National Academy of Engineering, a member of the Royal Irish Academy, a Fellow of the Institute of Electrical and Electronic Engineers and a foreign fellow of the Chinese Society for Electrical Engineering and has received two Fulbright Fellowships.

1- Hongbo Sun: Phase Management of Power Distribution Systems

Date and Location: Oct 1, 2024, 4pm at MIT 45-500A

Speaker: Hongbo Sun,Mitsubishi Electric Research Laboratories

Abstract: The power grid is designed as a three-phase alternating current (AC) system. To operate and control the grid effectively, it is crucial to know precisely how power sources and consumers are connected to its phases. Additionally, the three phases of the grid must be balanced as much as possible; otherwise, system stability, power quality, and equipment safety will be significantly compromised. This talk will cover two important topics for phase management in distribution systems: phase identification for loads without phase labels, and load phase reconfiguration for phase imbalance mitigation. First, we will present a data-driven method for phase identification, inspired by concepts from information theory. This method analyzes event signatures from customer locations and the feeder head, using data from smart meters. By examining the cross-entropy of their mutual information, it accurately identifies the correct phases for both wye and delta-connected loads. Next, we will introduce a data-driven approach for scheduling phase rebalancing over a finite horizon. Phase scheduling is formulated as a Markov Decision Process (MDP), and an Imitation Learning (IL) framework, implemented through Ensemble Random Forests, is used to streamline the process of sequential load phase swapping. The previously costly multi-objective optimization is now limited to steps where phase swapping is needed.

Speaker Bio: Dr. Hongbo Sun is currently a Senior Principal Research Scientist at Mitsubishi Electric Research Laboratories (MERL) in Cambridge, Massachusetts, USA. He received his Ph.D. degree in Electrical Engineering, specializing in Power Systems, from Chongqing University, China. His research areas include power transmission and distribution, power system operation and control, microgrids, hybrid AC/DC systems, and AI/ML applications.

Page updated

Google Sites

Report abuse