裂纹是导致塔架断裂的主要因素,而检测的裂纹缺陷信号中包含着丰富的时频信息,一般方法难以有效提取裂纹特征信息。针对这个问题,提出一种基于小波包最优基子带能量的裂纹特征提取方法。依据最小熵准则选取最优小波基及分解层数,并在该小波包基包含的所有分解系数中,提取能量系数较大的几个最优基子带构造特征向量,建立特征向量到缺陷状态的映射关系,进行裂纹缺陷的有效识别。并以某发射塔架为例,验证该方法对裂纹特征提取的有效性。结果表明,基于小波包最优基子带能量的特征提取方法,由于将小波包分解与能量谱相结合,充分利用检测信号的时—频域特征信息,因此可以有效识别结构的裂纹缺陷。同时也为进一步研究裂纹缺陷的定量问题提供依据。
Crack is the critical factor for launcher fracture.The defect signals are often localized both in time and frequency,so extract effective features form them by general transformation methods is very difficult.Owing to this problem,a method based on best basis sub bands energy of wavelet packet decomposition was proposed.The best basis of wavelet packet and optimal decomposition level were selected from wavelet packet library according to the entropy criterion.The wavelet packet basis contained many decompos...