摘要
提出了一种利用改进LCD算法提取模拟电路故障特征的方法。首先,针对LCD算法估计包络曲线时未考虑其凹凸性的问题,提出了一种改进的LCD算法并尽心了必要的随噪声影响仿真分析;其次,利用该算法对模拟电路被测信号进行分解,得到若干具有物理意义的分量;最后,使用基本尺度熵算法得到故障特征。运用Pearson相关性指标的仿真结果验证了该方法的有效性。
A new method of feature vector extraction in analog circuit based on improved LCD algorithm is proposed.Firstly,an improved LCD algorithm is proposed to solve the problem that the convexity of envelope curves is not considered in the standard algorithm.Secondly,the measured signal is decomposed by using thisalgorithm to obtain some signal components of physical significance.Finally,the fault features are obtained by using the basic scale entropy algorithm.The simulation results of Pearson correlation index show the effectiveness of this method.
作者
盛沛
许爱强
张加建
李恒
李辉
SHENG Pei;XU Ai-qiang;ZHANG Jia-jian;LI Heng;LI Hui(Naval Aeronautical University,Yantai 264001,China;292962 Force Support Department,Guangzhou 510320,China)
出处
《价值工程》
2019年第34期270-272,共3页
Value Engineering
基金
国家部委预研基金资助：基于残差向量的流体机械故障诊断与预研方法研究（9140A27020214JB1446）
关键词
信息观点
模拟电路
局部特征尺度分解
基本尺度熵
information theory view
analog circuit
local feature scale decomposition
basic scale entropy
