Machine Learning-based Service Restoration Scheme for Smart Distribution Systems with DGs and High Priority Loads

I. Kalysh, M. Kenzhina, N. Kaiyrbekov, H. S.V.S. Kumar Nunna, Aresh Dadlani, S. Doolla

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Reliability of power distribution systems is of very crucial concern due to cases of mass power outages that occur worldwide. Once an unscheduled outage takes place in power grids, the service restoration is triggered to rapidly return the system to normal conditions and minimize the severity of consequences. This paper proposes a self-healing power distribution grid restoration technique based on decentralized multi-agent systems with reinforcement learning. The system architecture is based on two types of zone agents: Inactive Zone Agent (IZA) and Active Zone Agent (AZA), where the IZA is activated provided that an agent is within the out-of-service area. This study contributes to the advancement of service restoration by endowing agents with learning ability. The reward computation proposed in this paper is based on the load priority factor, and also it ensures preserving the constraints within the limits. Case studies include a comparison of service restoration outcomes with load priority factor and DGs incorporated into the network. All simulations are implemented in the PowerWorld simulator for the medium voltage network of 11kV with 29 buses. The results of the study prove that embedding Q-learning algorithm into service restoration significantly improves the performance metrics and thus, increases the reliability of the distribution grids.

Original languageEnglish
Title of host publicationSEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728111568
DOIs
Publication statusPublished - Sep 1 2019
Event2nd International Conference on Smart Energy Systems and Technologies, SEST 2019 - Porto, Portugal
Duration: Sep 9 2019Sep 11 2019

Publication series

NameSEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies

Conference

Conference2nd International Conference on Smart Energy Systems and Technologies, SEST 2019
CountryPortugal
CityPorto
Period9/9/199/11/19

Fingerprint

Distribution System
Restoration
Learning systems
Machine Learning
Power Distribution
Grid
Outages
Q-learning
Reinforcement learning
Performance Metrics
Reinforcement Learning
System Architecture
Multi agent systems
Reward
Power System
Learning algorithms
Decentralized
Multi-agent Systems
Learning Algorithm
Simulator

Keywords

  • DG integration
  • multi-agent systems
  • Power distribution restoration
  • Q-learning
  • reinforcement learning

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Networks and Communications
  • Energy Engineering and Power Technology
  • Renewable Energy, Sustainability and the Environment
  • Electrical and Electronic Engineering
  • Control and Optimization

Cite this

Kalysh, I., Kenzhina, M., Kaiyrbekov, N., Kumar Nunna, H. S. V. S., Dadlani, A., & Doolla, S. (2019). Machine Learning-based Service Restoration Scheme for Smart Distribution Systems with DGs and High Priority Loads. In SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies [8849002] (SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SEST.2019.8849002

Machine Learning-based Service Restoration Scheme for Smart Distribution Systems with DGs and High Priority Loads. / Kalysh, I.; Kenzhina, M.; Kaiyrbekov, N.; Kumar Nunna, H. S.V.S.; Dadlani, Aresh; Doolla, S.

SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies. Institute of Electrical and Electronics Engineers Inc., 2019. 8849002 (SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kalysh, I, Kenzhina, M, Kaiyrbekov, N, Kumar Nunna, HSVS, Dadlani, A & Doolla, S 2019, Machine Learning-based Service Restoration Scheme for Smart Distribution Systems with DGs and High Priority Loads. in SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies., 8849002, SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies, Institute of Electrical and Electronics Engineers Inc., 2nd International Conference on Smart Energy Systems and Technologies, SEST 2019, Porto, Portugal, 9/9/19. https://doi.org/10.1109/SEST.2019.8849002
Kalysh I, Kenzhina M, Kaiyrbekov N, Kumar Nunna HSVS, Dadlani A, Doolla S. Machine Learning-based Service Restoration Scheme for Smart Distribution Systems with DGs and High Priority Loads. In SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies. Institute of Electrical and Electronics Engineers Inc. 2019. 8849002. (SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies). https://doi.org/10.1109/SEST.2019.8849002
Kalysh, I. ; Kenzhina, M. ; Kaiyrbekov, N. ; Kumar Nunna, H. S.V.S. ; Dadlani, Aresh ; Doolla, S. / Machine Learning-based Service Restoration Scheme for Smart Distribution Systems with DGs and High Priority Loads. SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies. Institute of Electrical and Electronics Engineers Inc., 2019. (SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies).
@inproceedings{447f3160060044cbb495e3708f0e4d85,
title = "Machine Learning-based Service Restoration Scheme for Smart Distribution Systems with DGs and High Priority Loads",
abstract = "Reliability of power distribution systems is of very crucial concern due to cases of mass power outages that occur worldwide. Once an unscheduled outage takes place in power grids, the service restoration is triggered to rapidly return the system to normal conditions and minimize the severity of consequences. This paper proposes a self-healing power distribution grid restoration technique based on decentralized multi-agent systems with reinforcement learning. The system architecture is based on two types of zone agents: Inactive Zone Agent (IZA) and Active Zone Agent (AZA), where the IZA is activated provided that an agent is within the out-of-service area. This study contributes to the advancement of service restoration by endowing agents with learning ability. The reward computation proposed in this paper is based on the load priority factor, and also it ensures preserving the constraints within the limits. Case studies include a comparison of service restoration outcomes with load priority factor and DGs incorporated into the network. All simulations are implemented in the PowerWorld simulator for the medium voltage network of 11kV with 29 buses. The results of the study prove that embedding Q-learning algorithm into service restoration significantly improves the performance metrics and thus, increases the reliability of the distribution grids.",
keywords = "DG integration, multi-agent systems, Power distribution restoration, Q-learning, reinforcement learning",
author = "I. Kalysh and M. Kenzhina and N. Kaiyrbekov and {Kumar Nunna}, {H. S.V.S.} and Aresh Dadlani and S. Doolla",
year = "2019",
month = "9",
day = "1",
doi = "10.1109/SEST.2019.8849002",
language = "English",
series = "SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
booktitle = "SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies",
address = "United States",

}

TY - GEN

T1 - Machine Learning-based Service Restoration Scheme for Smart Distribution Systems with DGs and High Priority Loads

AU - Kalysh, I.

AU - Kenzhina, M.

AU - Kaiyrbekov, N.

AU - Kumar Nunna, H. S.V.S.

AU - Dadlani, Aresh

AU - Doolla, S.

PY - 2019/9/1

Y1 - 2019/9/1

N2 - Reliability of power distribution systems is of very crucial concern due to cases of mass power outages that occur worldwide. Once an unscheduled outage takes place in power grids, the service restoration is triggered to rapidly return the system to normal conditions and minimize the severity of consequences. This paper proposes a self-healing power distribution grid restoration technique based on decentralized multi-agent systems with reinforcement learning. The system architecture is based on two types of zone agents: Inactive Zone Agent (IZA) and Active Zone Agent (AZA), where the IZA is activated provided that an agent is within the out-of-service area. This study contributes to the advancement of service restoration by endowing agents with learning ability. The reward computation proposed in this paper is based on the load priority factor, and also it ensures preserving the constraints within the limits. Case studies include a comparison of service restoration outcomes with load priority factor and DGs incorporated into the network. All simulations are implemented in the PowerWorld simulator for the medium voltage network of 11kV with 29 buses. The results of the study prove that embedding Q-learning algorithm into service restoration significantly improves the performance metrics and thus, increases the reliability of the distribution grids.

AB - Reliability of power distribution systems is of very crucial concern due to cases of mass power outages that occur worldwide. Once an unscheduled outage takes place in power grids, the service restoration is triggered to rapidly return the system to normal conditions and minimize the severity of consequences. This paper proposes a self-healing power distribution grid restoration technique based on decentralized multi-agent systems with reinforcement learning. The system architecture is based on two types of zone agents: Inactive Zone Agent (IZA) and Active Zone Agent (AZA), where the IZA is activated provided that an agent is within the out-of-service area. This study contributes to the advancement of service restoration by endowing agents with learning ability. The reward computation proposed in this paper is based on the load priority factor, and also it ensures preserving the constraints within the limits. Case studies include a comparison of service restoration outcomes with load priority factor and DGs incorporated into the network. All simulations are implemented in the PowerWorld simulator for the medium voltage network of 11kV with 29 buses. The results of the study prove that embedding Q-learning algorithm into service restoration significantly improves the performance metrics and thus, increases the reliability of the distribution grids.

KW - DG integration

KW - multi-agent systems

KW - Power distribution restoration

KW - Q-learning

KW - reinforcement learning

UR - http://www.scopus.com/inward/record.url?scp=85073363547&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85073363547&partnerID=8YFLogxK

U2 - 10.1109/SEST.2019.8849002

DO - 10.1109/SEST.2019.8849002

M3 - Conference contribution

AN - SCOPUS:85073363547

T3 - SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies

BT - SEST 2019 - 2nd International Conference on Smart Energy Systems and Technologies

PB - Institute of Electrical and Electronics Engineers Inc.

ER -