A digital converter station monitoring data fusion method for full-stop fault discrimination

As modern power systems evolve, high-voltage direct current (HVDC) technology has emerged as the preferred solution for transmitting power over long distances and large capacities. The performance of converter stations, which are central to HVDC systems, has a direct impact on the stability and reliability of the entire transmission network. Full-stop faults, where the converter station completely ceases operation, are particularly severe and can lead to widespread power outages, causing significant economic losses and social impacts. This study proposes a digital converter station monitoring data fusion method for full-stop fault discrimination. By integrating data from various sensors through Bayesian inference, the proposed method enhances the accuracy and reliability of fault detection. The model utilizes Convolutional Neural Networks (CNN), Long Short-Term Memory Networks (LSTM), and Attention mechanisms to process and analyze sensor data, demonstrating significant improvements in classification performance indicators such as accuracy, precision, recall, and F1 score after data fusion. The results underscore the critical role of data fusion in improving fault detection and ensuring the stable operation of power systems.