Research on the inversion method of pitting corrosion defects in buried pipelines based on weak magnetic detection

With the development of the oil and gas industry, the system construction of oil and gas pipeline networks is becoming increasingly perfect, and pipeline safety transportation has become a focus of attention. According to statistics, about a quarter of buried pipelines have been in use for more than 20 years. Under the influence of the surrounding soil environment, pitting corrosion defects often occur in buried pipelines, especially in the weld area. Weak magnetic detection technology, as a new type of non-destructive testing technology, has the advantages of easy operation, high sensitivity, and good adaptability, and is widely used in the field of buried oil and gas pipeline detection. At present, due to the dispersion of weak magnetic detection data and the scarcity of related sample data, the accuracy of defect inversion needs to be improved. In practical engineering, pitting defects are difficult to detect and difficult to detect, and current research has not involved predicting the angle of defects. In order to solve the above problems, this paper combines the Cubic Chaotic Mapping based Adaptive Whale Optimization Algorithm (CIWOA) based on CUBIC chaotic mapping and adaptive inertia weight with Back Propagation (BP) neural network to propose a weak magnetic detection based inversion method for buried pipeline pitting defects in terms of size and angle. Firstly, the influence of different pitting defects on the weak magnetic signal characteristics is analyzed through pipeline weak magnetic detection experiments, and inversion parameters are selected and a dataset is constructed. The Smote algorithm is used for data augmentation; Secondly, Kernel Principal Component Analysis (KPCA) is used for data dimensionality reduction to determine the principal components of the size and angle of pitting defects; Finally, establish the CIWOA-BP model for predicting the size and angle of pitting corrosion defects in buried pipelines. The results indicate that the CIWOA-BP model can accurately predict the size and angle of pitting defects in buried pipelines. Compared with other models, the CIWOA-BP model shows higher prediction accuracy in both pitting defect size and angle. Therefore, the inversion method of buried pipeline pitting defects based on weak magnetic detection can achieve accurate inversion of pitting defects, which has good application prospects for ensuring the safe transportation of buried pipelines.