논문명 |
딥러닝 기반 비침입 부하 모니터링 프레임워크에서의 LSTNet 응용/An Application of LSTNet in Deep Learning-based Non-intrusive Load Monitoring Framework |
저자명 |
김임규(Kim, Im-Gyu) ; 김현철(Kim, Hyun-Cheol) ; 신상용(Shin, Sang-Yong) |
발행사 |
한국건축친환경설비학회 |
수록사항 |
한국건축친환경설비학회 논문집
,
Vol.15 No.6
|
페이지 |
시작페이지(806) 총페이지(12)
|
ISSN |
1976-6483 |
주제분류 |
환경및설비
|
주제어 |
에너지 분해; 비침입 부하 모니터링(NILM); LSTNet; 스마트 그리드 Energy disaggregation; Non-Intrusive Load Monitoring (NILM); LSTNet; Smart Grid
|
요약2 |
In this study, a process of measuring and preprocessing power data for 4 types of home appliances and a deep training-based NILM technique were proposed. Active power of 4 types of home appliances (refrigerator, induction, TV, washer) was measured for about 3 weeks. The power data of each home appliance was measured to be aggregated in a smart meter. In order to disaggregate energy using LSTNet, four types of home appliances and smart meter power data were constructed as a training dataset. In the training process, we performed a parametric study to extract the optimal hyperparameter with the highest validation accuracy metric among the major parameters of LSTNet to verify the feasibility of NILM application of LSTNet. |
소장처 |
한국건축친환경설비학회 |