摘要
为提高电子装备故障预测结果的准确性,提出一种基于Bi-LSTM的电子装备故障预测方法。首先,在研究长短时记忆网络(LSTM)的基础上,提出将双向长短时记忆网络(Bi-LSTM)应用于电子装备故障预测,并给出基于Bi-LSTM的故障预测方法;其次,研究了模拟电路健康度的提取方法,并以某雷达带通滤波放大器的模拟电路为例开展基于仿真数据的故障预测研究;最后,将基于Bi-LSTM的故障预测方法应用于雷达发射机的三组状态监测数据中,开展基于实际数据的故障预测研究。通过电子装备仿真数据案例和实际数据案例的对比分析,说明基于Bi-LSTM的故障预测方法明显优于循环神经网络(RNN)和LSTM,能够提高电子装备故障预测的准确性。
In order to improve the accuracy of electronic equipment fault prediction results,a fault prediction method based on bi-directional long short term memory(Bi-LSTM) is proposed.Firstly,the research of long short term memory(LSTM) is carried out,and Bi-LSTM is proposed for fault prediction of electronic equipment,and a fault prediction method based on Bi-LSTM is given.Secondly,the extraction method of analog circuit health index is studied,and a fault prediction example based on the simulation data of a radar bandpass filter amplifier analog circuit is taken.Finally,the fault prediction method based on Bi-LSTM is applied to three groups of status monitoring data of radar transmitter to develop fault prediction research based on actual data.Through the comparative analysis of electronic equipment simulation data case and actual data case,it shows that the fault prediction method based on Bi-LSTM is obviously superior to the recurrent neural network(RNN) and LSTM,and Bi-LSTM can be used to improve the accuracy of electronic equipment fault prediction.
作者
倪祥龙
石长安
麻曰亮
刘磊
何健
Ni Xianglong;Shi Chang’an;Ma Yueliang;Liu Lei;He Jian(Unit 63891 of PLA,Luoyang 471003,China)
出处
《航空兵器》
CSCD
北大核心
2022年第6期102-110,共9页
Aero Weaponry
作者简介
倪祥龙(1988-),男,福建安溪人,博士。
