Study on spectral statistical characteristics of acoustic emission signals from the leakage of buried gas pipelines

Buried gas pipelines are more prone to leaks than aboveground gas pipelines, which are difficult to detect due to the harsh and hidden operating environment. To achieve factual dynamic monitoring of the operation status of buried gas pipelines in complex operation and maintenance environments, the analysis and research on the acoustic emission signals of pipeline operation based on the type of external perturbation and leakage conditions are conducted. The results show that (1) the maximum energy, energy variance and spectrum skewness of the signal decrease by 17.5 %, 22.5 % and 29.1 %, respectively, with the expansion of the leakage hole. However, the spectrum kurtosis increases by 40.5 % on average. The shape of leakage hole has great influence on acoustic emission signal. (2) Although the skewness coefficient of the maximum energy distribution increases, the spectrum skewness of the leakage signal is normally distributed and the distribution skewness coefficient is close to 0. The main frequency distribution in the leakage is scattered. (3) Under the action of disturbance, the decision tree model based on the spectrum distribution characteristics has a good effect on identifying the leakage state of buried gas pipelines, which has a good application prospect in the safe operation monitoring of gas pipelines.