Artificial Neural Network-based Small Signal Stability Analysis of Power Systems

The massive integration of renewable energy sources, due to their stochastic nature, is introducing unprecedented changes to power system dynamics. Near real-time small signal stability analysis can be crucial to identify the dominant modes and activate preventive control actions. However, model-based methods are time consuming for large scale systems and unfeasible for real-time operation.This paper proposes a black box approach for fast small signal stability analysis in power systems, based on artificial neural networks. A training set is generated from the results of modal analyses in a wide range of scenarios. Once trained, the neural network can predict, fast and accurately, the required stability indices. The effectiveness of the proposed method is tested on the WSCC 9-bus network.