王福忠, 乔珊珊, 田广强.基于长短期记忆网络的Vienna整流器故障预测[J].河南理工大学学报（自然科学版）,2023,42(3):111-117.

WANG F Z, QIAO S S, TIAN G Q.A fault prediction method of Vienna rectifier based on LSTM[J].Journal of Henan Polytechnic University(Natural Science) ,2023,42(3):111-117.

基于长短期记忆网络的Vienna整流器故障预测

王福忠1, 乔珊珊1, 田广强2

1.河南理工大学 电气工程与自动化学院，河南 焦作  454000;2.黄河交通学院 智能工程学院，河南 焦作  454950

摘要:为了掌握Vienna整流器的健康状态，提出基于长短期记忆（LSTM）网络的Vienna整流器故障预测模型。通过对电容和功率MOSFET的退化与故障特征分析，建立Vienna整流器整体电路性能和关键元器件退化之间的关系，选择输出电压变化值ω为整流器的故障特征参数。在此基础上，构建基于LSTM的Vienna整流器故障预测模型，采用Adam优化算法训练预测模型，实现对Vienna整流器故障特征参数的预测。仿真结果表明，该模型预测结果的均方根误差为0.123 3，平均绝对百分误差为0.101 8，该模型的预测精度较高，能够较好地实现Vienna整流器的故障预测。

关键词:Vienna整流器;长短期记忆网络;元器件退化;故障预测

doi:10.16186/j.cnki.1673-9787.2021070011

基金项目:国家重点研发计划专项项目（2016YFC0600906）；河南省科技攻关项目（212102210146）

收稿日期:2021/07/02

修回日期:2021/12/10

出版日期:2023/05/25

A fault prediction method of Vienna rectifier based on LSTM

WANG Fuzhong1, QIAO Shanshan1, TIAN Guangqiang2

1.School of Electrical Engineering and Automation，Henan Polytechnic University，Jiaozuo  454000，Henan，China;2.School of Intelligent Engineering，Huanghe Jiaotong University，Jiaozuo  454950，Henan，China

Abstract:In order to grasp the health status of the Vienna rectifier，a fault prediction model of the Vienna rectifier was proposed based on LSTM network.By analyzing the degradation and fault characteristics of capacitor and power MOSFET，a relationship was established between the circuit performance of the Vienna rectifier and the degradation of key components，so the output voltage variation value ω was selected as the fault characteristic parameter of the rectifier.On this basis，the fault prediction model of Vienna rectifier based on LSTM was constructed，and the Adam optimization algorithm was used to train the prediction model to realize the prediction of the characteristic parameters of the Vienna rectifier.The simulation results showed that，the RMSE of the prediction results of the model was 0.123 3，and the MAPE was 0.101 8.The prediction accuracy of the model was high，and the fault prediction of the Vienna rectifier could be achieved better.

Key words:Vienna rectifier;long and short-term memory network;component degradation;fault prediction

基于长短期记忆网络的Vienna整流器故障预测_王福忠.pdf