基于双向长短期记忆网络的高压输电线路短路故障识别方法Research on Short Circuit Fault Identification Method of High-Voltage Transmission Line Based on Bi-Directional Long Short-Term Memory
金乐
JIN Le;
摘要(Abstract)：
通过实时数字仿真器（RTDS）对多种工况下输电线路短路故障模型进行仿真，获得了故障辨识时所需的各种类型的故障信号。采用双向长短期记忆网络，利用采集的短路故障电流暂态量数据进行输电线路短路故障辨识。在利用双向长短期记忆网络辨识短路故障类型过程中，先将预处理后用于训练的5500种工况下输电线路双端短路电流暂态量时间序列传入网络输入层，再将学习到的数据传至隐含层进行训练，并不断对网络进行优化，得到训练模型。最后将整理后的测试样本集输入建立的模型中，对模型性能进行验证，短路故障辨识效果良好。
The short circuit fault model of transmission line under various working conditions is simulated by RTDS simulator, and the various types of fault signals required for fault identification are obtained. The BiLSTM (Bi-directional long Long Short-Term Memory) bidirectional short-term memory network is used and the collected short-circuit fault current transient data is used for short-circuit fault identification of transmission line respectively. In the process of identifying short-circuit fault types using bidirectional long short-term memory network, the two-terminal short-circuit current transient time series of transmission lines used for training under 5500 kinds of working conditions are firstly introduced into the network input layer, and then the learned data is passed transmitted to the implicit hidden layer for training, and the network is continuously optimized, obtaining and the training model is obtained. Finally, the collated test sample set is input into the established model to verify the performance of the model, and the identification effect is good.
关键词(KeyWords)：
输电线路;短路故障识别;双向长短期记忆网络;RTDS;故障信号
transmission line;short-circuit fault identification;bi-directional long short-term memory;RTDS;fault signal
基金项目(Foundation):
作者(Authors):
金乐
JIN Le;
DOI: 10.19929/j.cnki.nmgdljs.2021.0038
- 输电线路
- 短路故障识别
- 双向长短期记忆网络
- RTDS
- 故障信号
transmission line - short-circuit fault identification
- bi-directional long short-term memory
- RTDS
- fault signal