TY - GEN

T1 - An online HEMS scheduling method based on deep recurrent neural network

AU - Yoshida, Akira

AU - Fujimoto, Yu

AU - Amano, Yoshiharu

AU - Hayashi, Yasuhiro

N1 - Funding Information:
The part of this work is supported by JST CREST Gant Number JPMJCR15K5, and JSPS KAKENHI Grant Number JP18K14170.
Funding Information:
The part of this work is supported by KAKENHI Grant Number JP18K14170.
Publisher Copyright:
© ECOS 2019 - Proceedings of the 32nd International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems. All rights reserved.

PY - 2019

Y1 - 2019

N2 - For a daily-basis scheduling of an energy system, energy management system often enough to use not a global optimal scheduling but a near optimal scheduling. The article proposes an online scheduling framework without online optimization. The framework is built from two encoder-decoder architectures to extract features of time series; a multi-layer long short-term memory regression model for multi-step time-series forecasting, and multi-class and single-label classification model for on/off scheduling of a device. The models are estimated at offline, and return scheduling from historical time series as input, at online. We evaluate the accuracy of scheduling from the viewpoint of Kullback-Leibler divergence which measures the dissimilarity between two probability distributions. Through the numerical experiments, we demonstrate the usefulness of the proposed framework.

AB - For a daily-basis scheduling of an energy system, energy management system often enough to use not a global optimal scheduling but a near optimal scheduling. The article proposes an online scheduling framework without online optimization. The framework is built from two encoder-decoder architectures to extract features of time series; a multi-layer long short-term memory regression model for multi-step time-series forecasting, and multi-class and single-label classification model for on/off scheduling of a device. The models are estimated at offline, and return scheduling from historical time series as input, at online. We evaluate the accuracy of scheduling from the viewpoint of Kullback-Leibler divergence which measures the dissimilarity between two probability distributions. Through the numerical experiments, we demonstrate the usefulness of the proposed framework.

KW - Encoder-decoder

KW - Energy demand forecast

KW - Home energy management system

KW - Recurrent neural network

KW - Scheduling

UR - http://www.scopus.com/inward/record.url?scp=85079665663&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85079665663&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85079665663

T3 - ECOS 2019 - Proceedings of the 32nd International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems

SP - 1327

EP - 1335

BT - ECOS 2019 - Proceedings of the 32nd International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems

A2 - Stanek, Wojciech

A2 - Gladysz, Pawel

A2 - Werle, Sebastian

A2 - Adamczyk, Wojciech

PB - Institute of Thermal Technology

T2 - 32nd International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2019

Y2 - 23 June 2019 through 28 June 2019

ER -