摘要
为解决自适应最稀疏时频分析(ASTFA)方法中初始相位函数的选择问题,采用遗传算法(GA)对ASTFA的初始相位函数进行优化,提出了GA-ASTFA方法。进一步研究了GA-ASTFA方法抑制模态混淆的能力,分析结果表明,GA-ASTFA能较好地抑制模态混淆,分解得到的分量信号精度高,且可抑制分解中的伪分量。最后将GA-ASTFA方法用于转子碰摩故障诊断,实验分析结果表明,GA-ASTFA方法能有效提取转子碰摩故障特征信息。
To solve the choice of initial phase function, ASTFA method was proposed herein based on GA. The initial phase function of ASTFA was optimized by GA. The ability of avoiding mode mixing was studied. The analysis shows that the GA-ASTFA method has the ability of avoiding mode mixing and can suppress the pseudo components. And the component＇s accuracy decomposed via GA-ASTFA is better. Finally, the GA-ASTFA was applied to rub-impact fault diagnosis in rotor systems. Experimental analyses show that the GA-ASTFA method can extract the fault feature information effectively.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2016年第1期66-72,共7页
China Mechanical Engineering
基金
国家自然科学基金资助项目(51375152)
关键词
遗传算法
自适应最稀疏时频分析
经验模态分解
模态混淆
转子碰摩
genetic algorithm （GA）
adaptive and sparsest time-frequency analysis （ASTFA）
empirical mode decomposition（EMD）
mode mixing
rub-impact fault
