Detecting and Interpreting Faults in Vulnerable Power Grids With Machine Learning

Unscheduled power disturbances cause severe consequences both for customers and grid. To defend against such events, it is necessary to identify the causes of interruptions in the power network. In this work, we focus on the power grid of a Norwegian community in the Arctic experiences several faults whose sources are unknown. First, we construct a data set consisting of meteorological data and information about the current power quality logged by power-quality. Then, we adopt machine-learning techniques to predict the occurrence of faults. Experimental results that both linear and non-linear classifiers achieve good classification performance. This indicates that considered power quality and weather variables explain well the power disturbances. Interpreting the process of the classifiers provides valuable insights to understand the main causes of disturbances. features selection methods can only indicate which are the variables that, on average, mostly the fault occurrences in the dataset. Besides providing such a global interpretation, it is also important identify the specific set of variables that explain each individual fault. To address this challenge, we adopt recent technique to interpret the decision process of a deep learning model, called Integrated Gradients. The approach allows gaining detailed insights on the occurrence of a specific fault, which are valuable the distribution system operators to implement strategies to prevent and mitigate power disturbances.