Electric Device Abnormal Detection Based on IoT and Knowledge Graph

Abnormal detection is an essential task to the security and stability of power grid networking operation. Besides the status of power grid itself, the environment information is also necessary to the abnormal detection, such as temperature, humidity, dust and etc. Therefore, we apply the Internet of Things (IoT) technology to monitor the real time status of power grid and its related environment. Then a knowledge based method is proposed to find out the reference data for the abnormal detection and signal correlation algorithm is applied to calculate if the given IoT data is abnormal. The experiment results indicate that the proposed method can integrate the data from multiple resources and perform an accuracy analysis of the abnormality of the given IoT data. The overall accuracy improves from 64% with tradition threshold based method to 92% according to the test dataset.