Magnetic Flux Leakage Testing Method for Well Casing Based on Gaussian Kernel RBF Neural Network

摘要

Well casing integrity is important for the safe operations of oil wells, and is of great significance to detect well casing defects. Magnetic Flux Leakage (MFL) Detection Technology is widely used to detect the defects of various pipelines. Because the environment where well casing is laid in is usually very complicated, the system which based on magnetic flux leakage technology is not mature yet to detect well casing defects. The method of defects detection with RBF neural network based on Gaussian kernel is studied, by which parameters of well casing defects can be recognized. The training data samples were gathered from both the simulated data sets for 3-D finite element model and measured MFL data. Detection system suitable to casing inspection is established. The experiment result indicates that the system can detect the defect and identify its parameters effectively.