Power-grid stability predictions using transferable machine learning.

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.

Discontinuous phase transition in the Kuramoto model with asymmetric dynamic interaction.

We investigate the critical behavior of the modified Kuramoto model with an asymmetric dynamic interaction which has been proposed to explain the difference between the synchronized frequency and the average intrinsic frequency. We find that the discontinuous phase transition arises when oscillators interact only with other oscillators whose phases are ahead. From the comparison with the conventional Kuramoto model in which the interaction possesses phase reflection symmetry, we conclude that the dynamical symmetry breaking and the dynamic change in interaction structure play important roles in changing the transition nature from continuous to discontinuous ones.

Asymmetric dynamic interaction shifts synchronized frequency of coupled oscillators.

Interacting dynamic agents can often exhibit synchronization. It has been reported that the rhythm all agents agree on in the synchronized state could be different from the average of intrinsic rhythms of individual agents. Hinted by such a real-world behavior of the interaction-driven change of the average phase velocity, we propose a modified version of the Kuramoto model, in which the ith oscillator of the phase ϕi interacts with other oscillator j only when the phase difference [Formula: see text] - [Formula: see text] is in a limited range [-ßπ, απ]. From extensive numerical investigations, we conclude that the asymmetric dynamic interaction characterized by ß ≠ α leads to the shift of the synchronized frequency with respect to the original distribution of the intrinsic frequency. We also perform and report our computer-based synchronization experiment, which exhibits the expected shift of the synchronized frequency of human participants. In analogy to interacting runners, our result roughly suggests that agents tend to run faster if they are more influenced by runners ahead than behind. We verify the observation by using a simple model of interacting runners.