针对合金弹体的几种典型结构,采用超声检测技术进行缺陷检测,用不同尺寸的人工小孔和细槽代表实际的气泡和裂纹缺陷,研究了弹体内部缺陷的类型识别技术;通过对缺陷回波信号的总体经验模态分解,得到各个分量的能量分布和信息熵分布,并以此为特征量,采用支持向量机的方法进行了类型识别;用平均距离的方法对特征量进行评估和排序,将优化、精简个数后的特征量作为输入,在减少运算时间的同时,提高了缺陷识别率,最终的识别率为97.5%。
The ultrasonic testing method was adopted to detect the defect of the alloy projectile,and the type identification technology for the small hole and slot which are the alternatives of air-bubbles and cracks in the projectile was introduced.The ensemble empirical mode decomposition was used for extracting energy distribution and entropy distribution of IMFs.The method by evaluating and ordering the average distance of the features was used to reduce the run time and improve the identification rate.The identification rate is 97.5% for the above defects.