BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York BEGIN:DAYLIGHT DTSTART:20240310T030000 TZOFFSETFROM:-0500 TZOFFSETTO:-0400 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:EDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20231105T010000 TZOFFSETFROM:-0400 TZOFFSETTO:-0500 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:EST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20231113T150854Z UID:24CB098D-1652-4494-B002-F7D52F651505 DTSTART;TZID=America/New_York:20231110T122000 DTEND;TZID=America/New_York:20231110T132000 DESCRIPTION:Abstract:\n\nThe trend in electric power systems is the displac ement of traditional synchronous generation (e.g.\, coal\, natural gas) wi th renewable energy resources (e.g.\, wind\, solar photovoltaic) and batte ry energy storage. These energy resources require power electronic convert ers (PECs) to interconnect to the grid and have different response charact eristics and dynamic voltage and frequency stability issues compared to co nventional synchronous generators. As a result\, there is a need for next- generation methods to characterize and mitigate PEC-based dynamic stabilit y issues\, especially for converter-dominated power systems (e.g.\, island power systems\, remote microgrids).\n\nThis talk will discuss recent adva ncements in dynamic state estimation and control of battery energy storage systems. A framework will be introduced to provide fast frequency dynamic voltage support for converter-dominated power systems using both model-ba sed and model-free state estimation and control approaches. Model-based me thods will first be introduced using reduced-order power system dynamics e quations. Specifically\, a moving horizon state and model parameter estima tor provides dynamic state inputs to a model-predictive controller. These classic model-based methods are then compared to state-of-the-art model-fr ee methods from machine learning\; a neural ordinary differential equation s (NODEs)-based framework will be described to infer critical state inform ation of the power system frequency dynamics. The state information is use d by a soft-actor-critic (SAC) reinforcement learning-based controller. Th e model-based and model-free methods are compared for performance and comp utational efficiency. The topics presented will have broad applicability t o both undergraduate and graduate electrical and computer engineering stud ents\, including:\n\n- How is the global energy transition impacting elect ric power grid operations?\n- What is the interaction and future role of p ower electronics with the electric power system? and\n- What are the trade offs between complexity\, accuracy\, and computational tractability of tra ditional model-based and model-free machine learning approaches?\n\nSpeake r Bio:\n\nTimothy M. Hansen (IEEE Senior Member 2020) received the B.S. de gree in computer engineering with high honors from the Milwaukee School of Engineering\, Milwaukee\, WI\, USA\, in 2011\, and the Ph.D. degree in el ectrical engineering from Colorado State University\, Fort Collins\, CO\, USA\, in 2015. In 2014-2015\, he held a graduate research position in the Distributed Energy Systems Integration group at the National Renewable Ene rgy Laboratory\, Golden\, CO\, USA. He is currently the Harold C. Hohbach Endowed Associate Professor with the Electrical Engineering and Computer S cience Department\, South Dakota State University\, Brookings\, SD\, USA. His research interests are in the application of optimization\, high-perfo rmance/edge/quantum computing\, and distributed stochastic control to the areas of sustainable power and energy systems\, smart cities\, robotics\, and cyber-physical-social systems.\n\nSpeaker(s): Timothy M. Hansen\, \n\n Room: 336\, Bldg: ITE Building\, 371 Fairfield Way\, Storrs\, Connecticut\ , United States LOCATION:Room: 336\, Bldg: ITE Building\, 371 Fairfield Way\, Storrs\, Conn ecticut\, United States ORGANIZER:leila.chebbo@uconn.edu SEQUENCE:21 SUMMARY:Invited Talk: Dr. Timothy M. Hanses URL;VALUE=URI:https://events.vtools.ieee.org/m/382461 X-ALT-DESC:Description: <br /><p style="font-weight: 400\;"><strong><u>Abst ract:</u>&nbsp\;</strong></p>\n<p style="font-weight: 400\;">The trend in electric power systems is the displacement of traditional synchronous gene ration (e.g.\, coal\, natural gas) with renewable energy resources (e.g.\, wind\, solar photovoltaic) and battery energy storage. These energy resou rces require power electronic converters (PECs) to interconnect to the gri d and have different response characteristics and dynamic voltage and freq uency stability issues compared to conventional synchronous generators. As a result\, there is a need for next-generation methods to characterize an d mitigate PEC-based dynamic stability issues\, especially for converter-d ominated power systems (e.g.\, island power systems\, remote microgrids).< /p>\n<p style="font-weight: 400\;">This talk will discuss recent advanceme nts in dynamic state estimation and control of battery energy storage syst ems. A framework will be introduced to provide fast frequency dynamic volt age support for converter-dominated power systems using both model-based a nd model-free state estimation and control approaches. Model-based methods will first be introduced using reduced-order power system dynamics equati ons. Specifically\, a moving horizon state and model parameter estimator p rovides dynamic state inputs to a model-predictive controller. These class ic model-based methods are then compared to state-of-the-art model-free me thods from machine learning\; a neural ordinary differential equations (NO DEs)-based framework will be described to infer critical state information of the power system frequency dynamics. The state information is used by a soft-actor-critic (SAC) reinforcement learning-based controller. The mod el-based and model-free methods are compared for performance and computati onal efficiency. The topics presented will have broad applicability to bot h undergraduate and graduate electrical and computer engineering students\ , including:</p>\n<ul>\n<li style="font-weight: 400\;">How is the global e nergy transition impacting electric power grid operations?</li>\n<li style ="font-weight: 400\;">What is the interaction and future role of power ele ctronics with the electric power system? and</li>\n<li style="font-weight: 400\;">What are the tradeoffs between complexity\, accuracy\, and computa tional tractability of traditional model-based and model-free machine lear ning approaches?<strong>&nbsp\;</strong></li>\n</ul>\n<p style="font-weigh t: 400\;">&nbsp\;<strong><u>Speaker Bio:</u>&nbsp\;</strong></p>\n<p style ="font-weight: 400\;">Timothy M. Hansen (IEEE Senior Member 2020) received the B.S. degree in computer engineering with high honors from the Milwauk ee School of Engineering\, Milwaukee\, WI\, USA\, in 2011\, and the Ph.D. degree in electrical engineering from Colorado State University\, Fort Col lins\, CO\, USA\, in 2015. In 2014-2015\, he held a graduate research posi tion in the Distributed Energy Systems Integration group at the National R enewable Energy Laboratory\, Golden\, CO\, USA. He is currently the Harold C. Hohbach Endowed Associate Professor with the Electrical Engineering an d Computer Science Department\, South Dakota State University\, Brookings\ , SD\, USA. His research interests are in the application of optimization\ , high-performance/edge/quantum computing\, and distributed stochastic con trol to the areas of sustainable power and energy systems\, smart cities\, robotics\, and cyber-physical-social systems.&nbsp\;</p> END:VEVENT END:VCALENDAR