Stochastic model predictive control of residential fuel cell cogeneration unit with periodic forecast update for distributed architecture

Research output: Contribution to conferencePaper

Abstract

As communication standard, such as SEP 2.0 and ECHONET Lite, for digital communication of home appliances has been established, the research of networked control system is progressing in the domain of residential house. Instead of building new power station to fulfil new electricity requirement, such as electric vehicle charging, the concept, that an aggregator coordinates massive amount of home energy management system (HEMS) to supply negawatt power, is proposed. The aggregator requires to manipulate electricity consumption through massive amount of HMES as networked control system. As fundamental study for control algorithm of distributed architecture, this paper constructed one of the networked supervisory control system formulated as stochastic model predictive control scheme, and analysed primary energy consumption sensitivity against update interval of energy demand forecasting and operational strategy in addition to changing number of scenarios. The mainly results are concluded that: as the energy saving ratio in the case of 6 hours update interval has enough high point, which is 9.5% in comparison with the case of 24 hours update interval, the stochastic model predictive control scheme proposed in the case using six scenarios can downsample until 6 hours update interval.

Original languageEnglish
Publication statusPublished - 2017 Jan 1
Event30th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2017 - San Diego, United States
Duration: 2017 Jul 22017 Jul 6

Other

Other30th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2017
CountryUnited States
CitySan Diego
Period17/7/217/7/6

Fingerprint

cogeneration
Networked control systems
Model predictive control
Stochastic models
fuel cell
control system
Fuel cells
Electricity
Energy management systems
Domestic appliances
Communication
Electric vehicles
communication
electric vehicle
Energy conservation
Energy utilization
Control systems
power plant
electricity
forecast

Keywords

  • Distributed architecture
  • Fuel-cell cogeneration
  • Home EMS
  • Stochastic model predictive control
  • Update interval

ASJC Scopus subject areas

  • Environmental Science(all)
  • Energy(all)
  • Engineering(all)

Cite this

Yoshida, A., Yoshikawa, J., Fujimoto, Y., Amano, Y., & Hayashi, Y. (2017). Stochastic model predictive control of residential fuel cell cogeneration unit with periodic forecast update for distributed architecture. Paper presented at 30th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2017, San Diego, United States.

Stochastic model predictive control of residential fuel cell cogeneration unit with periodic forecast update for distributed architecture. / Yoshida, Akira; Yoshikawa, Jun; Fujimoto, Yu; Amano, Yoshiharu; Hayashi, Yasuhiro.

2017. Paper presented at 30th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2017, San Diego, United States.

Research output: Contribution to conferencePaper

Yoshida, A, Yoshikawa, J, Fujimoto, Y, Amano, Y & Hayashi, Y 2017, 'Stochastic model predictive control of residential fuel cell cogeneration unit with periodic forecast update for distributed architecture', Paper presented at 30th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2017, San Diego, United States, 17/7/2 - 17/7/6.
Yoshida A, Yoshikawa J, Fujimoto Y, Amano Y, Hayashi Y. Stochastic model predictive control of residential fuel cell cogeneration unit with periodic forecast update for distributed architecture. 2017. Paper presented at 30th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2017, San Diego, United States.
Yoshida, Akira ; Yoshikawa, Jun ; Fujimoto, Yu ; Amano, Yoshiharu ; Hayashi, Yasuhiro. / Stochastic model predictive control of residential fuel cell cogeneration unit with periodic forecast update for distributed architecture. Paper presented at 30th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2017, San Diego, United States.
@conference{6bbf818aa9bb41849223b76e81a4491e,
title = "Stochastic model predictive control of residential fuel cell cogeneration unit with periodic forecast update for distributed architecture",
abstract = "As communication standard, such as SEP 2.0 and ECHONET Lite, for digital communication of home appliances has been established, the research of networked control system is progressing in the domain of residential house. Instead of building new power station to fulfil new electricity requirement, such as electric vehicle charging, the concept, that an aggregator coordinates massive amount of home energy management system (HEMS) to supply negawatt power, is proposed. The aggregator requires to manipulate electricity consumption through massive amount of HMES as networked control system. As fundamental study for control algorithm of distributed architecture, this paper constructed one of the networked supervisory control system formulated as stochastic model predictive control scheme, and analysed primary energy consumption sensitivity against update interval of energy demand forecasting and operational strategy in addition to changing number of scenarios. The mainly results are concluded that: as the energy saving ratio in the case of 6 hours update interval has enough high point, which is 9.5{\%} in comparison with the case of 24 hours update interval, the stochastic model predictive control scheme proposed in the case using six scenarios can downsample until 6 hours update interval.",
keywords = "Distributed architecture, Fuel-cell cogeneration, Home EMS, Stochastic model predictive control, Update interval",
author = "Akira Yoshida and Jun Yoshikawa and Yu Fujimoto and Yoshiharu Amano and Yasuhiro Hayashi",
year = "2017",
month = "1",
day = "1",
language = "English",
note = "30th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2017 ; Conference date: 02-07-2017 Through 06-07-2017",

}

TY - CONF

T1 - Stochastic model predictive control of residential fuel cell cogeneration unit with periodic forecast update for distributed architecture

AU - Yoshida, Akira

AU - Yoshikawa, Jun

AU - Fujimoto, Yu

AU - Amano, Yoshiharu

AU - Hayashi, Yasuhiro

PY - 2017/1/1

Y1 - 2017/1/1

N2 - As communication standard, such as SEP 2.0 and ECHONET Lite, for digital communication of home appliances has been established, the research of networked control system is progressing in the domain of residential house. Instead of building new power station to fulfil new electricity requirement, such as electric vehicle charging, the concept, that an aggregator coordinates massive amount of home energy management system (HEMS) to supply negawatt power, is proposed. The aggregator requires to manipulate electricity consumption through massive amount of HMES as networked control system. As fundamental study for control algorithm of distributed architecture, this paper constructed one of the networked supervisory control system formulated as stochastic model predictive control scheme, and analysed primary energy consumption sensitivity against update interval of energy demand forecasting and operational strategy in addition to changing number of scenarios. The mainly results are concluded that: as the energy saving ratio in the case of 6 hours update interval has enough high point, which is 9.5% in comparison with the case of 24 hours update interval, the stochastic model predictive control scheme proposed in the case using six scenarios can downsample until 6 hours update interval.

AB - As communication standard, such as SEP 2.0 and ECHONET Lite, for digital communication of home appliances has been established, the research of networked control system is progressing in the domain of residential house. Instead of building new power station to fulfil new electricity requirement, such as electric vehicle charging, the concept, that an aggregator coordinates massive amount of home energy management system (HEMS) to supply negawatt power, is proposed. The aggregator requires to manipulate electricity consumption through massive amount of HMES as networked control system. As fundamental study for control algorithm of distributed architecture, this paper constructed one of the networked supervisory control system formulated as stochastic model predictive control scheme, and analysed primary energy consumption sensitivity against update interval of energy demand forecasting and operational strategy in addition to changing number of scenarios. The mainly results are concluded that: as the energy saving ratio in the case of 6 hours update interval has enough high point, which is 9.5% in comparison with the case of 24 hours update interval, the stochastic model predictive control scheme proposed in the case using six scenarios can downsample until 6 hours update interval.

KW - Distributed architecture

KW - Fuel-cell cogeneration

KW - Home EMS

KW - Stochastic model predictive control

KW - Update interval

UR - http://www.scopus.com/inward/record.url?scp=85048599574&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85048599574&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85048599574

ER -