Research Article
Early Warning Method of Abnormal Energy Consumption in Public Buildings Based on Multi-Level Analysis and Adaptive Weight
@INPROCEEDINGS{10.4108/eai.24-11-2023.2343367, author={Jiani Zeng and Longfei Ma and Baoqun Zhang and Xin Gao and Caie Hu and Siyue Lu and Hui Xu}, title={Early Warning Method of Abnormal Energy Consumption in Public Buildings Based on Multi-Level Analysis and Adaptive Weight }, proceedings={Proceedings of the International Conference on Industrial Design and Environmental Engineering, IDEE 2023, November 24--26, 2023, Zhengzhou, China}, publisher={EAI}, proceedings_a={IDEE}, year={2024}, month={2}, keywords={multi-level analysis; adaptive weight; public buildings; abnormal energy consumption; early warning method}, doi={10.4108/eai.24-11-2023.2343367} }- Jiani Zeng
Longfei Ma
Baoqun Zhang
Xin Gao
Caie Hu
Siyue Lu
Hui Xu
Year: 2024
Early Warning Method of Abnormal Energy Consumption in Public Buildings Based on Multi-Level Analysis and Adaptive Weight
IDEE
EAI
DOI: 10.4108/eai.24-11-2023.2343367
Abstract
In order to make the energy consumption of public buildings more reasonable, it is necessary to warn the changes of energy consumption. Therefore, an early warning method of abnormal energy consumption of public buildings based on multi-level analysis and adaptive weight is proposed. By analyzing the data variation between different levels, we can find out the existing abnormal data, optimize the abnormal data target and analyze the weight, so as to get the preprocessing results. Calculate the total energy consumption of public buildings in the life cycle, including the abnormal energy consumption of public buildings and the operating energy consumption; Input samples, use the minimum proportional distance to classify, output the weight results, get the output network early warning value, and get the final early warning result of abnormal energy consumption of public buildings. The experimental results show that this method can provide more stable and reliable early warning results, and the results are close to the actual energy consumption results.
