Research Article
High-Frequency Non-intrusive Load Monitoring System Based on KNN and QDA Ensemble Learning Algorithm
@INPROCEEDINGS{10.1007/978-3-031-60347-1_26, author={Wang Zihao and Zhou Zou}, title={High-Frequency Non-intrusive Load Monitoring System Based on KNN and QDA Ensemble Learning Algorithm}, proceedings={Mobile Multimedia Communications. 16th EAI International Conference, MobiMedia 2023, Guilin, China, July 22-24, 2023, Proceedings}, proceedings_a={MOBIMEDIA}, year={2024}, month={10}, keywords={Ensemble Learning NILM High-Frequency Dataset Electrical Inspection}, doi={10.1007/978-3-031-60347-1_26} }- Wang Zihao
Zhou Zou
Year: 2024
High-Frequency Non-intrusive Load Monitoring System Based on KNN and QDA Ensemble Learning Algorithm
MOBIMEDIA
Springer
DOI: 10.1007/978-3-031-60347-1_26
Abstract
In recent years Non-Intrusive Load Monitoring (NILM) technology has developed rapidly, with applications covering power management, smart homes, fault detection, equipment condition monitoring and other areas. In commercial building power inspection scenarios, traditional inspection methods often suffer from low efficiency, high costs, missed inspections and false inspections. At present, the use of NILM technology can identify the working status of each power equipment, significantly reducing the probability of missed and false detection. However, there is still the problem of not being able to balance classification efficiency and classification accuracy. In this study, a high frequency fast non-intrusive load monitoring system is designed. The use of high-frequency data sets significantly reduces the data acquisition time. To further reduce the time cost, an ensemble learning algorithm based on K-Nearest Neighbor (KNN) and Quadratic Discriminant Analysis (QDA) is proposed. The algorithm not only achieves an F1-score of 90%, but also has an average training time of 0.207 s. In comparison with six baseline algorithms such as stacked ensemble learning (SEL) and radial basis function neural networks (RBFNN), the proposed algorithm not only maintains a high F1-score, but also significantly reduces the average training time. In practical scenarios, the model can be deployed faster and maintain high recognition accuracy.
