Intelligent Monitoring of Power System Electricity Consumption Based on Isolation Forest Algorithm

Aiming at the problems of large amounts of power consumption data, high dimensionality, few samples of abnormal points in the actual data set, and high acquisition cost under the background of intelligent power distribution, an intelligent monitoring method based on fuzzy clustering and Isolation Fo...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2023 3rd International Conference on New Energy and Power Engineering (ICNEPE) S. 1051 - 1054
Hauptverfasser: Yin, Jianbing, Li, Fei, Xu, Xianghai, Cheng, Ying, Zhu, Lei, An, Dongsheng, Ma, Xianghui, Liu, Shasha, Liu, Da
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 24.11.2023
Schlagworte:
Behavioral sciences
Classification algorithms
Clustering algorithms
Costs
Electric power system
Forestry
iForest algorithm
Intelligent monitoring
Monitoring
Power demand
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Aiming at the problems of large amounts of power consumption data, high dimensionality, few samples of abnormal points in the actual data set, and high acquisition cost under the background of intelligent power distribution, an intelligent monitoring method based on fuzzy clustering and Isolation Forest is proposed. Firstly, the historical electricity consumption data is preprocessed, and the power users with the same electricity consumption behavior are classified by the fuzzy clustering algorithm, and the users with different electricity consumption behaviors are intelligently monitored and analyzed by the iForest algorithm. The results show that, under the same data set, the iForest algorithm has higher precision and recall than the LOF and K-means algorithms in intelligent monitoring of power consumption data. It is proved that the method proposed in this paper is more suitable for intelligent monitoring and analysis with fewer abnormal points in electricity consumption data.
DOI:10.1109/ICNEPE60694.2023.10429584