基于GAF-CNN的电力系统暂态稳定评估

Power System Transient Stability Assessment Based on GAF-CNN

摘要: 为保障电力系统安全稳定运行，针对电力系统暂态稳定评估（TSA）问题，提出了一种基于数据图像化的深度学习方法。首先，通过格拉姆角场（GAF）将原始的电力系统数据转为易于区分稳定与失稳的二维图像。其次，利用得到的二维图像数据集训练卷积神经网络（CNN）模型并进行在线应用。最后，通过在CEPRI 36节点系统和含风机的IEEE39节点系统、IEEE300节点系统中对所提TSA方法进行验证，结果表明了所提方法的有效性。

Abstract: To ensure the safe and stable operation of power systems,the paper proposes a deep learning method based on data visualization for the power system transient stability assessment（TSA）problem. Firstly,the raw power system data is converted into two-dimensional images that can easily distinguish between stability and instability by Gramian angular field（GAF）. Then an obtained two-dimension image collection is used to train a convolutional neural network（CNN）model and the online application is done.Finally,the proposed method is verified in the CEPRI 36-BUS system,IEEE 39-BUS system with wind turbine and IEEE 300-BUS system,and the results show the effectiveness of the proposed method.