Study of Gas Pipeline Leak Detection Based on Hilbert Marginal Spectrum

Gas pipeline leakage will lead to great economic losses. So, the study of leak detection on gas pipelines is very important. A leak detection method based on Hilbert-Huang transform (HHT) has been proposed. First, the signal is transformed via HHT, than the Hilbert marginal spectrum will be acquired, which can reflect changing regularity of the signal amplitude. Through compare the marginal spectrum curve between normal condition and leak condition, then the character frequency band of the signal will be acquired. The energy in the character frequency band is treated as the feature vector. Markov distance between the test data and sample data is established, and it is used to judge the working conditions of the gas pipeline. The validity of this leak detection approach has been assessed via experiment. According to the difference of the leak arrival time for both the upstream and downstream and sound speed in the gas of the pipeline, we can get the localization of the leak point.