Intelligent Transient Stability Assessment for High-Renewable Sending-End System with VSC-HVDC Links

With the increasing use of voltage source converter-based high-voltage direct current (VSC-HVDC) lines for large-scale renewable energy transmission over long distances, intricate stability issues in high-renewable sending-end systems have also gradually emerged. One of the critical challenges in the practical operation of the sending-end systems is how to efficiently and reliably assess its transient stability against large disturbances. To address this challenge, this paper proposes an intelligent transient stability assessment (TSA) method based on convolutional neural networks (CNN). Unlike traditional model-based methods that rely on equivalent models or time-domain simulations, this approach leverages CNN to learn system stability features from various transient scenarios represented by simulation data. Specifically, taking both steady-state information and transient responsive time-series data as the inputs, it automatically extracts and learns the key stability features for system transient stability margin estimation. Compared with conventional model-based solutions, the proposed method can significantly improve the efficiency and accuracy of transient stability margin assessment for complex sending-end power systems. Numerical tests on a realistic high-renewable sending-end system with VSC-HVDC links demonstrate the efficacy of the proposed method.