Convolutional neural network‐based power system frequency security assessment

Weak inertia characteristics of power systems with high penetrations renewables have become a prominent problem for frequency security. To solve this problem, convolutional neural network (CNN)-based deep learning approach is applied to realize rapid security assessment (FSA). First, the time series feature autonomously mined from wide-area measurement data serve as input data. By doing so, complex construction process quantity avoided. A structure then used establish non-linear mapping relationship between features and indicators end-to-end system prediction. Next, evaluation accuracy proposed optimized by tuning key parameters in CNN-based model. Through error analysis wind penetration sensitivity study, anti-interference performance model demonstrated. Finally, effectiveness FSA verified case studies modified 16-machine 68-node China Southern Power Grid.