Research on Leak Detection and Localization Algorithm for Oil and Gas Pipelines Using Wavelet Denoising Integrated with Long Short-Term Memory (LSTM)–Transformer Models
Ma, Y.; Shang, Z.; Zheng, J.; Zhang, Y.; Weng, G.; Zhao, S.; Bi, C. Research on Leak Detection and Localization Algorithm for Oil and Gas Pipelines Using Wavelet Denoising Integrated with Long Short-Term Memory (LSTM)–Transformer Models. Sensors 2025, 25, 2411. https://doi.org/10.3390/s25082411
The goal of this study was to develop a more accurate leak detection and localization algorithm for oil and gas pipelines by integrating wavelet denoising with a Long Short-Term Memory (LSTM)–Transformer hybrid model.
Key findings showed that the proposed model, by combining LSTM's temporal pattern recognition and the Transformer’s attention mechanism, achieved a leak prediction accuracy of 99.995% and a leak localization error margin below 2.5%. Wavelet denoising effectively filtered noise from pressure signals, enhancing the model’s ability to detect small leaks (>0.6% of flow) without generating false alarms.
Although Fluxim's tools like Paios, Setfos, and Laoss were not used in this specific study, similar advanced data acquisition and signal processing tools are critical in comparable experimental setups involving real-time sensor data, high-resolution pressure monitoring, and model validation — highlighting the importance of precise, reliable characterization equipment in complex pipeline diagnostics.
The study is significant because it offers a robust solution to overcome traditional challenges in pipeline monitoring, such as poor generalization, heavy reliance on manual data interpretation, and high false alarm rates. It demonstrates the value of combining advanced signal processing with AI models to significantly improve infrastructure safety and efficiency.