Power-grid stability predictions using transferable machine learning

Abstract

Complex network analyses have provided clues to improve power-grid stability with the help of numerical models. The high computational cost of numerical simulations, however, has inhibited the approach, especially when it deals with the dynamic properties of power grids such as frequency synchronization. In this study, we investigate machine learning techniques to estimate the stability of power-grid synchronization. We test three different machine learning algorithms—random forest, support vector machine, and artificial neural network—training them with two different types of synthetic power grids consisting of homogeneous and heterogeneous input-power distribution, respectively. We find that the three machine learning models better predict the synchronization stability of power-grid nodes when they are trained with the heterogeneous input-power distribution rather than the homogeneous one. With the real-world power grids of Great Britain, Spain, France, and Germany, we also demonstrate that the machine learning algorithms trained on synthetic power grids are transferable to the stability prediction of the real-world power grids, which implies the prospective applicability of machine learning techniques on power-grid studies.

Toward dynamic stability assessment of power grid topologies using graph neural networks

“…There are a number of applications dealing with GNNs and different power flow-related tasks [24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39] and to predict transient dynamics in microgrids. 40 There is also literature using conventional ML methods dealing with the basin stability 41,42 in the context of power grids. Nauck et al 43 use small GNNs to predict the dynamic stability on small datasets.…”

An efficient smart grid stability prediction system based on machine learning and deep learning fusion model

“…The authors in [19] applied various ML techniques for smartgrid stability prediction and achived good results. They used ML models to predict power-grid synchronization stability [20]. The algorithms used in experiemnts are RF, SVM and artifical neural networks (ANN).…”

Predicting the power grid frequency of European islands

“…Fully modelling the short-term frequency dynamics is very challenging due to its combined stochastic-deterministic nature and the numerous external influences [10,17,18], suggesting the use of data-driven approaches [19] and the usage of machine learning to predict time series as well as to understand and control energy systems [20][21][22][23]. These data-driven approaches complement modelling-based approaches [10,24] as they do not make any assumptions about the governing dynamical equations but still provide forecasts and explanations of the system.…”