Research on Fault Detection for Ring Network Cabinet

In order to ensure the safe and stable operation of the power supply system, an online fault detection method based on local features of data is proposed for the loop network cabinet data modeling and online monitoring. Using the strategy of local feature extraction based on neighborhood preserving embedding (NPE) algorithm, real-time data features are obtained through multiple measurement variable information and environment variable information of ring network cabinet, and a fault detection model of ring network cabinet based on data features is constructed. The constructed NPE model is applied to the on-line detection of real ring network cabinet, and the original data space is divided into irrelevant feature space and data residual space. According to these two spaces, the monitoring statistics of Hotelling T 2 and the sum of the squared prediction errors (SPE) are constructed respectively, and based on these two monitoring statistics, the online real-time monitoring and fault alarm of ring network cabinet are realized. This method is applied to the fault detection of ring network cabinet, and the test results prove the effectiveness of this method in fault detection of ring network cabinet.