A Neural Network-based Application for Oil and Gas Pipeline Defect Depth Estimation

Experienced engineers utilize Magnetic Flux Leakage (MFL) sensors to scan oil and gas pipelines for the purpose of localizing and sizing different defect types. The huge amount of raw data obtained by these sensors, however, makes the inspection task error-prone and timeconsuming. In this paper, we propose a defect depth estimation approach using artificial neural networks of various architectures. Discriminant features, which characterize different defect depth patterns, are first obtained from the raw data. Neural networks are then trained using these features. The Levenberg-Marquardt back-propagation learning algorithm is adopted in the training process, during which the weight and bias parameters of the networks are tuned to optimize their performances. Compared with the performance of pipeline inspection techniques reported by service providers such as GE and ROSEN, the results obtained using the method we proposed are promising. For instance, within ±10% errortolerance range, the obtained estimation accuracy is 86%, compared to only 80% reported by GE; and within ±15% error-tolerance range, the achieved estimation accuracy is 89% compared to 80% reported by ROSEN.