Abstract：Considering the key techniques and difficulties of defect visualization in high-speed MFL testing, to preprocesses defect MFL signals through lifting wavelet package method, a theoretic analysis to the influence for MFL signals of the velocity of the sensor in high-speed MFL testing based on electromagnetic field essential theory is given. Present RBF neural network compensation measure which based on regularization theory, and gained the velocity invariance response of the defect MFL signals through net mapping. And RBF networks which optimized by ant colony algorithm for two-dimensional inversion of defect MFL signals is applied to implements Profile Reconstruction of 2D defect.
［1］Jin-feng Ding,Yi-hua Kang,Xin-jun Wu. Tubing Thread Inspection by Magnetic Flux Leakage［J］. NDT & E International, 2006, 39(1): 53-56.
［2］Haueisen J,Unger R,Beuker T, et al. Evaluation of inverse algorithms in the analysis of magnetic flux leakage data ［J］. IEEE Transactions On Magnetics, 2002, 38(3): 1481-1488.
［3］Park G S, Park S H. Analysis of the Velocity-Induced Eddy Current in MFL Type NDT［J］. IEEE Transactions on Magnetics, 2004, 40 （2）:663-666.
［4］杨理践,马凤铭,高松巍.管漏磁在线检测系统的研究［J］.仪器仪表学报(增刊),2004,25(4):1052-1054.
［5］Sweldens W. The Lifting Scheme:A Construction of Second Generation Wavelets［J］. SIAM J Math Anal,1997,29(2):511-546.
［6］徐章隧,徐英,王建斌,等.裂纹漏磁定量检测原理与应用［M］.北京:国防工业出版社,2005,152-156.
［7］Mandayam S,Udpa L,Udpa S S,et al. Invariance Transformations for Magnetic Flux Leakage Signals［J］. IEEE Transactions on Magnetics,1996,32（3）:1577-1580.
［8］Mandayam S. Invariance Algorithms for Nondestructive Evaluation［J］. IEEE, 2001, 6:3397-3400.
［9］周琛, 马峥. 基于模糊聚类与非线性最小二乘法的径向基网络设计［J］. 系统仿真学报, 2002, 14(7): 859-862.
［10］武方方,赵银亮.一种基于蚁群聚类的径向基神经网络［J］.西安交通大学学报,2006,40(4):386-389.
2011, Vol. 32 
