Detecting and Interpreting Faults in Vulnerable Power Grids With Machine Learning
Matteo Chiesa, Filippo Maria Bianchi, Odin Foldvik Eikeland, Inga Setså Holmstrand, Sigurd Bakkejord
Abstract
Unscheduled power disturbances cause severe consequences both for customers and grid\noperators. To defend against such events, it is necessary to identify the causes of interruptions in the power\ndistribution network. In this work, we focus on the power grid of a Norwegian community in the Arctic\nthat experiences several faults whose sources are unknown. First, we construct a data set consisting of\nrelevant meteorological data and information about the current power quality logged by power-quality\nmeters. Then, we adopt machine-learning techniques to predict the occurrence of faults. Experimental results\nshow that both linear and non-linear classifiers achieve good classification performance. This indicates that\nthe considered power quality and weather variables explain well the power disturbances. Interpreting the\ndecision process of the classifiers provides valuable insights to understand the main causes of disturbances.\nTraditional features selection methods can only indicate which are the variables that, on average, mostly\nexplain the fault occurrences in the dataset. Besides providing such a global interpretation, it is also important\nto identify the specific set of variables that explain each individual fault. To address this challenge, we adopt\na recent technique to interpret the decision process of a deep learning model, called Integrated Gradients. The\nproposed approach allows gaining detailed insights on the occurrence of a specific fault, which are valuable\nfor the distribution system operators to implement strategies to prevent and mitigate power disturbances.