论文部分内容阅读
为有效去除兰姆波检测信号中的冗余信息和识别多个模态,应用匹配追踪方法对兰姆波信号进行处理。在Chirplet原子基础上添加弯曲算子进行改进,以更好地匹配频散和多模式兰姆波信号的特征。由改进的Chirplet原子组成过完备字典,使用基于遗传算法的匹配追踪(GAMP)信号稀疏分解方法,从过完备字典中选出与待分析信号相匹配的最佳原子,利用最佳匹配原子和对应的分解系数进行信号重构和时频分析。研究结果表明,改进后的Chirplet原子更能反映出兰姆波信号的非线性时频变化特征,得到的时频分布与频散曲线的弯曲特性能很好的吻合。采用改进后的Chirplet原子匹配追踪方法可以获取更加精确的走时信息,为后续兰姆波损伤定位成像奠定基础。
In order to effectively remove the redundant information in the Lamb wave detection signal and identify multiple modes, the matching tracking method is used to process the Lamb wave signal. Bending operators are added to the Chirplet atom for improvement to better match the characteristics of dispersion and multimode Lamb waves. A perfect dictionary is composed of modified Chirplet atoms. Genetic algorithm based GAMP signal sparse decomposition method is used to select the best atom matching the signal to be analyzed from the overcomplete dictionary. By using the best matched atom and the corresponding The decomposition factor for signal reconstruction and time-frequency analysis. The results show that the modified Chirplet atoms can better reflect the nonlinear time-frequency characteristics of the Lamb wave signal. The obtained time-frequency distribution can be well fitted to the bending characteristics of the dispersion curve. The improved Chirplet atomic matching tracking method can obtain more accurate travel time information, which lays the foundation for subsequent Lamb wave damage localization imaging.