Proactive Semi-Supervised Machine Learning Method for Stability Estimation in Smart Grids

Various types of machine learning methods are being used to predict and estimate power stability in smart grids. These include both supervised and unsupervised machine learning methods. This paper proposes the use of a semi-supervised learning method using a graph model for proactive learning in smart grids. The proposed method starts by labeling the unlabeled bus voltage data using information from the partially labeled bus voltage data. Then, it evaluates the score of the labeling and predicts the system stability by using correlation matrices. The method is applied for five power system test cases including the IEEE 14 bus, 30 bus, 39 bus, 57 bus, and 118 bus systems. The results confirm a high degree of correlation between the predicted results and the actual values of the terminal voltages of all the nodes in the experimental systems. The method can be used to predict the effects of disturbances in smart grids and related systems of systems.