摘要
我军部分舰船装备中含有大量可调元件,可通过人为调节的方式使关键参数达到指标要求。为此,设计一套有针对性的故障特征提取算法及故障分类方法,并用Labview完成程序的开发。该软件用虚拟示波器完成信号采集,用局部特征尺度分解及分形盒维数算法完成故障特征提取,用概率神经网络完成故障类别的判定。最后,以某电路典型故障状态为例,运用该软件进行故障类别判断。结果表明,该软件能有效应对电路中多点连续超差情况下的故障状态识别。
There are a lot of adjustable components in the equipment of some ships of our army,which can make the key parameters meet the requirements of the indexes by means of artificial adjustment.To this end,a set of targeted fault feature extraction algorithm and fault classification method is designed,and Labview is used to complete the development of the program.The software uses virtual oscilloscope to complete signal acquisition,local characteristic scale decomposition and fractal box dimension algorithm to complete fault feature extraction,and probabilistic neural network is used to complete fault classification determination.Finally,taking the typical fault state of a circuit as an example,this software is used to judge the fault category.The results show that the software can effectively deal with fault state recognition in the case of multi-point continuous out-of-tolerance.
作者
盛沛
王庆江
陈黎明
SHENG Pei;WANG Qingjiang;CHEN Liming(Naval Aviation University,Yantai 264001)
出处
《舰船电子工程》
2021年第9期140-143,共4页
Ship Electronic Engineering
基金
国家部委预研基金项目(编号:9140A27020214JB1446)资助。
关键词
故障诊断
模拟电路
虚拟示波器
fault diagnosis
analog circuit
virtual oscilloscope