Anomaly Recognition Method for Massive Data of Power Internet of Things Based on Bayesian Belief Network

Abstract

At this stage, the massive data of the power Internet of Things (IoT) is collected and stored in a centralized way, this leads to the inefficient use and analysis of a large amount of information in the power IoT. The anomaly identification of massive data in the power IoT is the guarantee to effectively judge the operation status and security status of the power IoT. Therefore, this project takes the power grid IoT as the research object, and based on Bayesian theory, studies a new data anomaly identification method. First, the data conversion method is used to standardize the massive data of the power IoT, and the discrete wavelet transform method is used to preprocess the massive data of the power IoT. Secondly, the clustering algorithm is used to determine the feature density index of the massive data of the power IoT, and the high-order statistics method is used to extract the abnormal features of the massive data of the power IoT. Finally, based on Bayesian belief network, on this basis, establish and train the big data anomaly identification model of the power grid IoT to realize the big data anomaly identification of the power grid. The experimental results show that the proposed method can effectively improve the accuracy of anomaly recognition of the massive data of the power IoT.