Premium
Prediction model on disturbance of maintenance operation during real‐time pricing adaptive control for building air‐conditioners
Author(s) -
Matsukawa Shun,
Takehara Masanori,
Otsu Hideyuki,
Morikawa Junji,
Inaba Takashi,
Kondo Seiji,
Ninagawa Chuzo
Publication year - 2019
Publication title -
ieej transactions on electrical and electronic engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.254
H-Index - 30
eISSN - 1931-4981
pISSN - 1931-4973
DOI - 10.1002/tee.22921
Subject(s) - air conditioning , control (management) , computer science , artificial neural network , power (physics) , smart grid , term (time) , power consumption , control engineering , reliability engineering , engineering , artificial intelligence , electrical engineering , mechanical engineering , physics , quantum mechanics
It is desirable to predict the occurrence of disturbance during power consumption management such as the Real‐Time Pricing (RTP) adaptive control in the future smart grid. Maintenance operation is one of the most power‐consuming operations of multitype air‐conditioners in office buildings. Because the operation occurs stochastically and abruptly, it is extremely difficult to predict the occurrence timing from the RTP control system's point of view. In this research, we propose a prediction model that forecasts the impending occurrence of the operation. The model was implemented as a Long Short‐Term Memory (LSTM) neural network because the occurrence timing depends on the air‐conditioner's operation history. The prediction accuracy was evaluated and then the contribution of the prediction model to the RTP control goodness score was estimated. In our example case, even though prediction accuracy was not perfect, the improvement of the score was estimated by up to 21%. © 2019 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.