Application of XGBoost in Identification of Power Quality Disturbance Source of Steady-state Disturbance Events

Abstract

In the era of IoT, time-series pattern recognition is not a trivial task considering their fast dynamically changing characteristic. The distribution and trend changes of these time-series data such as current or voltage could reflect emerging environment event. In response to the increasing power quality disturbances in the power grid, the XGBoost algorithm are used to identify multiple sources of power quality. Firstly, statistical methods are used to extract features from power quality disturbance sources. These features are computed through different statistic method with aims to reflect the time-series distribution. Secondly, a training data set is constructed and a XGBoost classifier is trained based on the generated training data sets. Furthermore, the prior knowledge of some interference sources is added on this basis, and then it is applied to power quality interference source identification. Experimental results show that this method can effectively identify power quality disturbance sources, and the proposed method has good robustness and noise immunity.