High-Reliability Supply Path Search Method Based on Bidirectional Long Short-Term Memory (Bi-LSTM) Networks and Dijkstra’s Algorithm

Abstract:
With the increasing demand of key users for power supply reliability, it has become a hot research topic to make full use of and effectively integrate all kinds of multi-source heterogeneous data to provide the power supply path with optimal reliability under different power supply guarantee scenarios. This paper first carries out principal component analysis on the data contents such as the state of power grid components and the operation mode of the external environment and extracts the data features from various collection platforms. Then, based on the bidirectional long and short-term memory (Bi-LSTM) network, the environmental scene information and power grid situation information are correlated with their corresponding fault results, and the component failure rate calculation model taking into account the scene is built. Finally, the grid topology is transformed into a weighted directed graph, the Dijkstra algorithm is adopted to comprehensively consider the failure rate of each power supply path, and in this regard, the required high-reliability path is searched. The IEEE 30-node simulation example shows that the method studied in this paper is effective in searching for the most reliable power supply path for important users.