Behavioral Threat Assessment of Power Monitoring Systems Based on Machine Learning

An evaluation test method based on machine learning has been created to address the power monitoring system’s lack of internal operation behavior detection assessment.By employing this technique, problems with the power monitoring system’s mixed operation behavior, the operation role’s ambiguity, and inaccurate evaluation of operation behavior have been resolved. To pre -classify operational behavior, deepen the potential correlations between different types of information, extract behavioral traits from different operational roles to build a person image, and then train LOF, iForest and K-means anomaly directive detection models using the idea of bagging in integrated learning to evaluate operational behavior are all steps in the method. The method, which has been experimentally verified, works well for assessing operational behavior in a power monitoring system.