東華大學圖書館 |

Towards a Resilient and Intelligent Energy Management System Design for Distribution Networks with High Renewable Energy Penetration.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Towards a Resilient and Intelligent Energy Management System Design for Distribution Networks with High Renewable Energy Penetration./
作者:	Shirsat, Ashwin.
面頁冊數:	1 online resource (286 pages)
附註:	Source: Dissertations Abstracts International, Volume: 84-04, Section: B.
Contained By:	Dissertations Abstracts International84-04B.
標題:	Deep learning. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=29342798click for full text (PQDT)
ISBN:	9798351499307

Towards a Resilient and Intelligent Energy Management System Design for Distribution Networks with High Renewable Energy Penetration.
Shirsat, Ashwin.

Towards a Resilient and Intelligent Energy Management System Design for Distribution Networks with High Renewable Energy Penetration. - 1 online resource (286 pages)

Source: Dissertations Abstracts International, Volume: 84-04, Section: B.

Thesis (Ph.D.)--North Carolina State University, 2022.

Includes bibliographical references

With rapidly plummeting costs of renewable distributed generation and their enabling-technologies such as energy storage, the integration of highly uncertain and non-dispatchable power generation resources into the grid continues to rise. This trend is further accelerated by the necessity of adopting green-energy sources to curb carbon emissions to combat climate change. Although having distributed generation scattered throughout the network provides a path to improve network resiliency and achieve energy independence goals, the volatility of such systems makes it difficult for existing energy management systems to operate the grid efficiently and reliably. Furthermore, unlike traditional generation resources that can be dispatched instantaneously, renewable generation lacks much-needed controllability. Hence, the level of uncertainty under which energy management systems have to take optimal scheduling and dispatching decisions, is significantly high.Disruptive events such as extreme weather events and targetted cyber attacks significantly impact the incumbent power distribution infrastructure, leading to extended-duration outages and loss of critical services. Traditionally seen as high-impact, low-probability events, their occurrence frequency is rising due to the exacerbating effects of climate change and growing tensions among different nation-states. This calls for the need to enhance grid resiliency to prepare for, adapt to, and absorb the impacts of the extreme events and maintain a reliable supply of electricity. Multiple avenues are available today to enhance grid resiliency by integrating distributed generation and through new innovative grid technologies.This research aims to design novel intelligent energy management systems for the resilient operation of electricity networks with increased PV and energy storage penetration; ranging from small residential-level networks to large-scale distribution grids. The goal is to design an optimal generation-dispatch and load-curtailment decision-making strategies under extreme uncertainty stemming from the volatile nature of renewable generation, electricity demand, and unforeseeable conditions caused by extreme events. A key emphasis is placed on designing algorithms that have the following features: are computationally efficient, are spatiotemporally generalizable, take uncertainty-aware decisions, prioritize critical loads, maintain reliable electricity supply for extended outage durations, provide reliable grid-forming support to distributed generators, and take decisions that can be safely dispatched on real-world systems. We divide this research into three parts. In the first part, an intelligent self-learning energy management system for residential electric systems with rooftop PV generation and local energy storage is proposed to achieve the goal of energy independence. The system is controlled using a stochastic model predictive control (SMPC) approach. A bivariate Markov chain-based solution is proposed for computationally efficient uncertainty modeling. Finally, an unsupervised recursive Bayesian approach with observational aging is proposed for self-adapting the uncertainty model on the fly to reflect the dynamic temporal operating conditions. The second part of this research proposes a novel, secure, and adaptive hierarchical multi-timescale (SA-HMTS) framework for the resilient operation of active distribution networks under extended duration outages. The framework is modular, which means that it can be implemented for individual microgrids (MGs) or for distribution network integrated MGs to jointly support the MG and the distribution grid. In addition, a novel approach called Delayed Recourse is developed to shield the proposed framework from extreme forecast error scenarios and adapt the decisions to mitigate any operational downsides caused by forecast errors. The framework is then tested in a hardware-in-loop environment to validate its efficacy in a simulated dynamic environment mimicking a real-world network.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2023

Mode of access: World Wide Web

ISBN: 9798351499307Subjects--Topical Terms:

3554982
Deep learning.
Index Terms--Genre/Form:

542853
Electronic books.

Towards a Resilient and Intelligent Energy Management System Design for Distribution Networks with High Renewable Energy Penetration.
LDR:05551nmm a2200421K 4500 001 2355045
005 20230515064555.5
006 m o d
007 cr mn ---uuuuu
008 241011s2022 xx obm 000 0 eng d
020 $a 9798351499307
035 $a (MiAaPQ)AAI29342798
035 $a (MiAaPQ)NCState_Univ18402039742
035 $a AAI29342798
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Shirsat, Ashwin. $3 3695433
245 1 0 $a Towards a Resilient and Intelligent Energy Management System Design for Distribution Networks with High Renewable Energy Penetration.
264 0 $c 2022
300 $a 1 online resource (286 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 84-04, Section: B.
500 $a Advisor: Johnson, Jeremiah ; Baran, Mesut ; Lu, Ning ; Tang, Wenyuan.
502 $a Thesis (Ph.D.)--North Carolina State University, 2022.
504 $a Includes bibliographical references
520 $a With rapidly plummeting costs of renewable distributed generation and their enabling-technologies such as energy storage, the integration of highly uncertain and non-dispatchable power generation resources into the grid continues to rise. This trend is further accelerated by the necessity of adopting green-energy sources to curb carbon emissions to combat climate change. Although having distributed generation scattered throughout the network provides a path to improve network resiliency and achieve energy independence goals, the volatility of such systems makes it difficult for existing energy management systems to operate the grid efficiently and reliably. Furthermore, unlike traditional generation resources that can be dispatched instantaneously, renewable generation lacks much-needed controllability. Hence, the level of uncertainty under which energy management systems have to take optimal scheduling and dispatching decisions, is significantly high.Disruptive events such as extreme weather events and targetted cyber attacks significantly impact the incumbent power distribution infrastructure, leading to extended-duration outages and loss of critical services. Traditionally seen as high-impact, low-probability events, their occurrence frequency is rising due to the exacerbating effects of climate change and growing tensions among different nation-states. This calls for the need to enhance grid resiliency to prepare for, adapt to, and absorb the impacts of the extreme events and maintain a reliable supply of electricity. Multiple avenues are available today to enhance grid resiliency by integrating distributed generation and through new innovative grid technologies.This research aims to design novel intelligent energy management systems for the resilient operation of electricity networks with increased PV and energy storage penetration; ranging from small residential-level networks to large-scale distribution grids. The goal is to design an optimal generation-dispatch and load-curtailment decision-making strategies under extreme uncertainty stemming from the volatile nature of renewable generation, electricity demand, and unforeseeable conditions caused by extreme events. A key emphasis is placed on designing algorithms that have the following features: are computationally efficient, are spatiotemporally generalizable, take uncertainty-aware decisions, prioritize critical loads, maintain reliable electricity supply for extended outage durations, provide reliable grid-forming support to distributed generators, and take decisions that can be safely dispatched on real-world systems. We divide this research into three parts. In the first part, an intelligent self-learning energy management system for residential electric systems with rooftop PV generation and local energy storage is proposed to achieve the goal of energy independence. The system is controlled using a stochastic model predictive control (SMPC) approach. A bivariate Markov chain-based solution is proposed for computationally efficient uncertainty modeling. Finally, an unsupervised recursive Bayesian approach with observational aging is proposed for self-adapting the uncertainty model on the fly to reflect the dynamic temporal operating conditions. The second part of this research proposes a novel, secure, and adaptive hierarchical multi-timescale (SA-HMTS) framework for the resilient operation of active distribution networks under extended duration outages. The framework is modular, which means that it can be implemented for individual microgrids (MGs) or for distribution network integrated MGs to jointly support the MG and the distribution grid. In addition, a novel approach called Delayed Recourse is developed to shield the proposed framework from extreme forecast error scenarios and adapt the decisions to mitigate any operational downsides caused by forecast errors. The framework is then tested in a hardware-in-loop environment to validate its efficacy in a simulated dynamic environment mimicking a real-world network.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2023
538 $a Mode of access: World Wide Web
650 4 $a Deep learning. $3 3554982
650 4 $a Failure analysis. $3 3563948
650 4 $a Communication. $3 524709
650 4 $a Optimization techniques. $3 3681622
650 4 $a Systems design. $3 3433840
650 4 $a Neural networks. $3 677449
650 4 $a Decision making. $3 517204
650 4 $a Energy management. $3 3680718
650 4 $a Alternative energy sources. $3 3561089
650 4 $a Performance evaluation. $3 3562292
650 4 $a Data exchange. $3 3685752
650 4 $a Distance learning. $3 3557921
650 4 $a Markov analysis. $3 3562906
650 4 $a Alternative energy. $3 3436775
650 4 $a Design. $3 518875
650 4 $a Educational technology. $3 517670
650 4 $a Energy. $3 876794
650 4 $a Finance. $3 542899
655 7 $a Electronic books. $2 lcsh $3 542853
690 $a 0459
690 $a 0363
690 $a 0800
690 $a 0389
690 $a 0710
690 $a 0791
690 $a 0508
690 $a 0454
690 $a 0796
710 2 $a ProQuest Information and Learning Co. $3 783688
710 2 $a North Carolina State University. $3 1018772
773 0 $t Dissertations Abstracts International $g 84-04B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=29342798 $z click for full text (PQDT)