Research on Anomaly Detection Method for Virtual Power Plant Measurement Data Based on Isolation Forest Algorithm

Due to the complexity and diversity of virtual power plants, there are often outliers in the measurement data, which may have a negative impact on the state estimation, optimization scheduling, and decision support of the power system. Therefore, this paper proposes an anomaly detection method based on the Isolation Forest algorithm. Clean and normalize the measurement data of the virtual power plant. Introducing the Isolation Forest algorithm, combined with the characteristics and requirements of virtual power plant measurement data, the dataset is randomly divided to construct multiple isolation trees, and the degree of data anomaly is determined based on the path length from data points to the root node. The experimental results show that this method can accurately identify most outliers while maintaining good recognition ability for normal data.