Abstract:

The substation fault reasoning rules and analysis applications summarized by human are incomplete，difficult and vulnerable to interfering signals. Fault reasoning's configurations are of low reusability and have to take time sequence of input signals into consideration， which cannot be effectively solved by traditional machine learning algorithms. Thus， a substation intelligent fault reasoning method based on long-short-term memory recurrent neural network （LSTM-RNN） and natural language processing（NLP） technology is proposed. Based on the analysis on the application scenarios of fault reasoning， the overall architecture and key technologies of the intelligent fault reasoning method are expounded. The feasibility of the intelligent fault reasoning method that does not rely on man-made rules is verified by the data of application tests，and the LSTM-RNN works better than other machine learning algorithms in the scenarios that time sequence of signals can be memorized.

Key words: substation, intellectualization, fault reasoning, LSTM, NLP