Dynamic reinforcement identification method for weak links in distribution network based on deep graph neural network

Abstract Aiming at the insufficient recognition accuracy of weak points in the distribution network, a dynamic enhancement recognition method for weak nodes in the distribution network based on a deep graph neural network is proposed. It integrates the complex topology of the distribution network, automatically extracts key features, and accurately locates and reinforces weak links in real time. The results show that the recognition accuracy and system stability are significantly improved.