فهرست مطالب
Contents\n1 Introduction\n 1.1 Outline\n 1.2 Influence Factors of Magnetic Flux Leakage Testing\n 1.2.1 The Influence of Magnetizing\n 1.2.2 The Influence of Testing\n 1.3 Research Status of Defect Quantification Method of Magnetic Flux Leakage Testing\n 1.3.1 MFL Detection Theory and Technology Development\n 1.3.2 Research on Quantitative Inversion Technique of Defects\n 1.3.3 The Problems of Defect Quantitative Inversion in MFL Testing\n2 Testing Signal Processing Method\n 2.1 Data Acquisition and Storage\n 2.1.1 Data Acquisition of Magnetic Flux Leakage Testing\n 2.1.2 Organization and Storage of Data\n 2.2 Data Compression and Noise Reduction Method\n 2.2.1 Testing Data Compression\n 2.2.2 Noise Reduction Method of the Testing Signal\n3 Quantitative Method of Magnetic Flux Leakage Testing\n 3.1 Outline\n 3.2 Defect Quantification Method based on Statistical Identification\n 3.2.1 Pretreatment of Magnetic Flux Leakage Signal\n 3.2.2 Definition and Extraction of the Waveform Features\n 3.2.3 Statistical Identification of Defect Length\n 3.2.4 Multivariate Statistical Analysis Method\n 3.2.5 Statistical Identification of Defect Width\n 3.2.6 Statistical Identification of Defect Depth\n 3.3 Defect Quantification Method Based on Radial Basis Function Neural Network\n 3.3.1 The Iterative Method Based on Neural Network\n 3.4 Defect Quantification Method Based on 3D Finite Element Neural Network\n 3.4.1 Discretization Principle of Finite Element Method\n 3.4.2 Finite Element Neural Network\n 3.4.3 From One Dimension to Three Dimensions\n 3.4.4 Solving Positive and Inverse Problems Using FENN\n 3.4.5 Analysis on the Advantages of the FENN\n 3.4.6 The Optimization of FENN\n4 Defect Profile Inversion of Three-Dimensional MFL Detection\n 4.1 The Characteristics of Three-Dimensional MFL Signal\n 4.1.1 The Basic Characteristics of the Signal\n 4.1.2 The Signal Changes with the Size of Defects\n 4.2 Random Search Iterative Inversion Method of Defect 3D Contour\n 4.2.1 The Region Segmentation and Recognition\n 4.2.2 Defect Opening Contour Detection Method\n 4.2.3 Defect 3D Profile Mesh Model\n 4.2.4 Random Searching Iterative Inversion for the 3D Profile of Defects\n 4.3 Iterative Inversion of Neural Networks for Defect 3D Profiles\n 4.3.1 Main Feature Extraction of 3D MFL Detection Signal\n 4.3.2 Defect 3D Profile Strip Model\n 4.3.3 Forward Prediction of MFL Signal Based on RBF Neural Network\n 4.3.4 Iterative Inversion of RBF Neural Network for 3D Profile of Defects\n 4.4 Multistage Successive Approximation Inversion Method for Defect 3D Profile\n 4.4.1 Multistage Inversion of Defect Profiles\n 4.4.2 Progressive Refinement of the Defect Mesh Model\n 4.4.3 Progressive Refinement of Division Size of Finite Element Model\n 4.4.4 Three-Dimensional Profile Inversion Test of Actual Defect\n 4.5 Influence and Correction of Sampling Precision on the 3D Profile Inversion of Defects\n 4.5.1 The Influence of Sampling Precision\n 4.5.2 Interpolation Correction Method for Three-Dimensional MFL Signal\n 4.5.3 Test Verification\n5 Three-Dimensional MFL Imaging Detection\n 5.1 Characteristics of Three-Dimensional MFL Signal\n 5.1.1 Parameter Definition of Pits\n 5.1.2 Parameter Definition of Horizontal Grooves\n 5.1.3 Parameter Definition of Tangential Grooves\n 5.2 Defect Classification Quantitative Method under Complete MFL Signal\n 5.2.1 Defect Classification Method Based on RBF Neural Network\n 5.2.2 Defect Quantization Method Based on BP Neural Network\n 5.3 Quantization and Display Method of Defect Under Incomplete Signal\n 5.3.1 Defect Edge Recognition\n 5.3.2 Defect Depth Estimation\n 5.3.3 Real-Time Display of Defect Under Incomplete Signal\nReferences\nIndex
