The Fault Location Method for Power Communication Networks Based on the Bi-GRU Network

In recent years, the topology of power communication networks has become increasingly complex, resulting in diverse and intricate alarm data. Traditional correlation algorithms struggle with fault localization and identification, proving inefficient and inadequate for the timely processing of the large volume of network disturbance alarms in power communication networks. The continuous development of artificial intelligence offers new approaches and robust methods for addressing fault localization issues in power communication networks. This paper proposes a neural network model based on a Bidirectional Gated Recurrent Unit (Bi-GRU) with an attention mechanism. The model captures long-term dependencies in time series data using the Bi-GRU and employs the attention mechanism to highlight critical time points. This enables direct identification of root alarms, which are the sources of faults in the communication network, thus improving the accuracy and efficiency of fault localization. Experimental results demonstrate that the proposed algorithm exhibits superior performance in terms of fault localization accuracy and response time through learning and training on the alarm correlation dataset in power communication networks.