摘要
以磁为激励源进行轧辊表面及内部裂纹缺陷的检测,基于磁荷理论的裂纹数学模型的建立和神经网络的分段循环逼近的裂纹信号定量分析,实现了裂纹无损定量检测。经在铁磁材料上模拟实验取得了满意效果。
The internal and surface cracks in rolls were detected with magnet as exciting source. Quantitative detection of cracks was realized by the establishment of mathematical model according to magneton theory and quantitative analysis of crack signals based on the piecewise cyclic approximation of neural network. Simulative test on ferromagnetic material was satisfactory.
出处
《无损检测》
1999年第3期121-123,125,共4页
Nondestructive Testing
基金
国家自然科学基金资助项目
关键词
漏磁探伤机
裂纹
轧辊
检测装置
Magnetic leakage inspection machine Crack Mathematical model Signal processing
