Leak Detection of Underwater Pipelines Based on Machine Learning Algorithms

Due to the complex marine environment, the leak detection of undersea natural gas pipelines is facing great difficulties. To solve the problem, this paper takes the bubble flow generated by the pipeline leakage as the detection object and uses the ultrasonic transmission method to conduct nondestructive detection. The wavelet threshold method is used to denoise the recorded signals, and the two bubble flow detection models based on the support vector regression (SVR) and BP neural network are proposed. The experimental results and the proposed evaluation indicators indicate that the support vector regression algorithm performs better, which is more suitable for the detection of underwater bubble flow. Therefore, the machine learning algorithm provides effective help for the development of submarine pipeline detection technology in a more efficient and accurate direction.