Radiation noise identification of gas leaks from underwater structures based on higher-order spectral features

Underwater gas-tight structures will generate radiated noise when gas leaks, the identification of the radiation noise can be used to detect the presence of leaks in underwater structures and to remedy them at an early stage. The generation of gas leakage radiation noise is a very complex nonlinear process, so it is more difficult to make an accurate description and identification of it. An experimental environment is designed for the identification of gas leakage radiation noise from underwater structures, and a support vector machine identification method based on higher-order spectral features is proposed. Firstly, by extracting the higher-order spectrum of the signal, the signal features are made more prominent, thus improving the classification accuracy of the target recognition algorithm, and then the signal is classified based on the support vector machine to identify the gas leakage. The research results show that the method can effectively classify and identify simulated leakage targets, and has good prospects for engineering applications.