In response to the issue of massive data volume generated by magnetic flux leakage (MFL) non-destructive testing in oil and gas pipelines, an intelligent data compression method based on a targeted one-dimensional fully convolutional autoencoder network is proposed. Firstly, a data preprocessing module is designed to generate high-quality data required for subsequent processing, taking into account the characteristics of MFL data. Secondly, a data block classification algorithm is developed to calculate peak values for segmented differential data, and based on a predefined targeted threshold, distinguish different types of MFL data. Subsequently, based on the distinct data types, targeted one-dimensional fully convolutional autoencoder models are constructed to effectively achieve dimensionality reduction compression and reconstruction of the MFL data. Through practical experimental analysis, the reconstruction error such as MAE is reduced by about 27.7% and the compression ratio is improved by about 14% compared with traditional methods such as PCA. In addition, compared with ID-AE, the proposed 1D-FCAE reduces 206.8 k, 1.58G, and 80 s in parameters, memory usage, and training time, respectively, and reduces compression and decompression time by 60 ms and 69 ms, respectively, validating that it is easy to be applied in industrial environments with limited resources.

With the robust growth of the national economy, the demand for oil and natural gas continues to rise, heightening the significance of pipeline transportation in the energy sector. However, long-term pipeline operations are often subjected to factors such as corrosion, aging, and damage, which can result in leaks and safety incidents, posing significant threats to life, property, and environmental integrity. Consequently, timely and precise detection of pipeline defects and estimation of their sizes hold paramount practical importance. This research endeavors to employ advanced information technology and artificial intelligence to explore methods for pipeline defect detection and size estimation grounded in prior knowledge. The aim is to enhance the accuracy and reliability of pipeline defect diagnosis and ensure the safe operation of pipelines. The primary innovative work includes the development of a preprocessing method based on prior knowledge, the design of an adaptive algorithm for estimation of defect size, and the creation of an algorithm for estimation of deep cascade pipeline defect size. These methods effectively combine traditional mechanisms and data-driven approaches, leveraging the strengths of both to improve performance, accuracy, and robustness. The proposed methodology demonstrates superior accuracy and stability in defect inversion, providing strong technical support for the quantitative assessment of pipeline defects, which is significant for fault diagnosis and the precise maintenance of long-distance pipelines.