APP下载

新非平稳信号分析方法-自适应径向墨西哥草帽核时频分布

2015-12-30李志农,朱明,龙盛蓉

振动与冲击 2015年10期
关键词:故障诊断

第一作者李志农男,博士,教授,1966年8月生

新非平稳信号分析方法—自适应径向墨西哥草帽核时频分布

李志农,朱明,龙盛蓉(南昌航空大学无损检测技术教育部重点实验室, 南昌330063)

摘要:据自适应最优核设计准则,构造新的自适应径向墨西哥草帽型核函数时频分析方法。该方法能据分析信号分布自适应调节核函数的扩展方向及宽度,使在信号自项方向尽可能延伸,而不在信号自项方向的互项尽可能抑制,能克服传统时频分布中固定核函数缺乏对信号自适应能力的不足。给出自适应径向墨西哥草帽型核函数时频分布定义及算法,并与传统的短时傅里叶变换、Wigner-Ville分布及小波变换方法对比分析。仿真结果表明,所提方法优于传统的时频分布方法,能更有效处理非平稳信号,可获得较传统时频分布分辨率及抗噪更高性能。通过用于转子裂纹故障诊断实验结果表明,该方法能有效反映裂纹故障的严重程度。

关键词:径向墨西哥草帽核;自适应时频分布;自适应优化;故障诊断;裂纹转子

基金项目:国家自然科学基金(51075372,50775208, 51265039);江西省教育厅科技计划项目(GJJ12405);湖南科技大学机械设备健康维护湖南省重点实验室开放基金(201204) ;江西省研究生创新基金项目(YC2013-S214)

收稿日期:2014-02-08修改稿收到日期:2014-05-23

中图分类号:TN911;TH165.3;TH17;TP391文献标志码:A

A new method for non-stationary signal analysis-adaptive radial mexican-hat kernel time-frequency distribution

LIZhi-nong,ZHUMing,LONGSheng-rong(MOE Key Laboratory of Nondestructive Testing, Nanchang Aeronautical University, Nanchang 330063, China)

Abstract:A new adaptive radial Mexican-hat kernel (RMK) time-frequency distribution method was proposed according to the design criteria of adaptive optimal kernel. The characteristic of the proposed method was that RMK could self-adaptively adjust the expansion direction and width of the kernel function according to the distribution of the analyzed signal. The kernel function expanded as far as possible in auto-term direction, and was suppressed as far as possible in cross-term direction to overcome the deficiency of lack of self-adaptability for the fixed kernel function in the traditional time-frequency distribution. Here, the definition and the algorithm of RMK were presented, and the proposed method was compared with the traditional time-frequency distribution, such as, short Fourier transformation, Wigner-Ville distribution and wavelet transformation. The simulation results showed that the proposed method is superior to the traditional time-frequency distribution, it can more effectively be used to process non-stationary signals, and obtain the higher time-frequency resolution and anti-interference performance. Finally, the proposed method was applied in the fault diagnosis of rotor crack. The test results showed that the proposed method is very effective and can reflect the severity of rotor crack fault.

Key words:radial mexican-hat kernel; adaptive time-frequency distribution; adaptive optimization; fault diagnosis; cracked rotor

工程中所得动态信号往往具有非平稳性。由于非平稳动态信号的统计特性与时间有关,故对其须同时进行时、频分析[1]。时频分析能提供频谱内容随时间变化的信息,为非平稳信号分析的有力工具[2]。……

登录APP查看全文

猜你喜欢

故障诊断
基于包络解调原理的低转速滚动轴承故障诊断
ILWT-EEMD数据处理的ELM滚动轴承故障诊断
冻干机常见故障诊断与维修
基于EWT-SVDP的旋转机械故障诊断
数控机床电气系统的故障诊断与维修
基于改进的G-SVS LMS 与冗余提升小波的滚动轴承故障诊断
改进的奇异值分解在轴承故障诊断中的应用
基于LCD和排列熵的滚动轴承故障诊断
基于KPCA和PSOSVM的异步电机故障诊断
基于WPD-HHT的滚动轴承故障诊断