Metric Mutation Anomaly Detection Method of the Power Grid Dispatching Automatic System Based on Machine Learning and Statistics

Abstract With the increase of the scale, function and complexity of power grid dispatching automatic system, the difficulty of system operation and maintenance increases significantly. This paper presents a metric mutation anomaly detection method based on the combination of machine learning and statistical algorithm. First, the system data is smoothed by machine learning algorithm, and the difference between the smoothed value and the real value is calculated. Then, the difference was used to detect data anomaly by ensemble of statistical method. The effectiveness and accuracy of the method is verified by the test data of the power grid dispatching automatic system. The method can quickly locate the anomaly components of the system and assist the system maintenance personnel in making decisions.