부산대학교 도서관

The current intelligent power equipment alarm correlation analysis technology is mainly used for linkage inference through threshold limits to achieve correlation analysis of power equipment alarms, which leads to more alarm omissions due to the lack of identification of power equipment faults. In this regard, a research on the correlation analysis method of power equipment alarms based on machine learning is proposed. The support vector machine algorithm is used to reduce the dimensionality of equipment fault data and identify the key fault data features of power equipment. The alarm information of power equipment is collected and stored, and an alarm correlation analysis model is established. In the experiments, the proposed alarm correlation analysis method is verified for alarm accuracy. The experimental analysis results show that the alarm correlation analysis model constructed by the proposed method has a low false alarm rate and high alarm analysis accuracy.