基于金属磁记忆法的焊接缺陷特征值提取

Feature Extraction of Welding Defects Based on Metal Magnetic Memory

摘要: 研究了磁记忆检测技术在焊缝检测中的应用情况。采集了有缺陷和无缺陷试件分别在无应力作用下和加载后的磁记忆信号, 并进行对比分析。以小波包变换为理论基础, 对磁记忆信号进行小波包分解和小波包能量谱特征分析。试验证明, 采用磁记忆信号的能量集中位置来判断缺陷的区域是可行的。

Abstract: The application of metal magnetic memory testing(MMMT) in weld inspection was investigated. Through experiments, MMMT signals of two samples of being stress free and load imposed were gathered respectively, and comparative analysis was carried out. On the theoretical basis of wavelet packet transform, MMMT signals were processed by wavelet packet decomposition and wavelet energy spectrum analysis. The feasibility of judging the area of defect by the position of energy concentration was proved.