摘要
由于在漏磁场正问题求解、信号反演等方面还没有形成系统的理论和方法,因此漏磁检测信号的定量解释一直是无损检测技术领域的研究重点。在综述国内外漏磁信号定量解释方法研究现状的基础上,分析了由漏磁信号定量描述缺陷特征的技术特点以及模式匹配法、统计分析法的局限性,重点探讨了利用人工神经网络方法解释漏磁信号的优点和不足,并指出了可视化、多传感器信息融合等漏磁信号定量解释技术的研究发展方向。
Because it is still difficult to find an efficient theory and method systemically in the solutions to electromagnetic fields and its inverse problems, the signal quantitative interpretation methods have been the key to magnetic flux leakage testing technique all the time. Based on the review of current research and development for MFL signal interpretation methods, the difficulty of defect quantitative characterization via MFL testing signals was analyzed, and the capabilities and limitations of the model matching method and the statistical relation model were presented. The advantages and disadvantages of the artificial neural network were mainly discussed. And such research tendencies of the MFL signals interpretation as visualization, multi-sensors data fusion were pointed out.
出处
《无损检测》
北大核心
2007年第7期407-411,426,共6页
Nondestructive Testing
基金
国家自然科学基金资助项目(50305017)
中国博士后科学基金资助项目(2005038358)
湖北省教育厅青年人才基金项目资助(2007A098)
关键词
漏磁检测
定量解释
人工神经网络
Magnetic flux leakage testing
Quantitative interpretation
Artificial neural network
