Study On The Leakage Detection System And Related Algorithm Of Urban Underground Gas Pipeline Network

As a clean and efficient energy, nature gas gains increasingly gratifying attention around the world. And as the main method of gas resource transportation, the pipeline's security would be much more important. Especially for the accident of the cities underground gas pipeline, it will not only lead to huge property losses, but also have potentially dangerous to cause huge casualties. As the rapid development of city gas pipeline network, an accurate and rapid detection method is imperative.There are lots of methods for the gas pipeline leakage detection which are commercially available on the market, but few of them are suitable for urban gas pipeline. According to the national urban gas pipeline network laying standards, an experimental platform is built in an idle factory to simulate the buried pipeline leakage. With this platform, this paper applies acoustic emission, wavelet, higher-order statistics and generalized cross correlation to the leakage detection and location algorithm research. This paper's main contents are as follows:1. According to national standards, a 36m pipeline, which is 30m with 6m over the elbow, was laid for the simulation of the actual operation of the gas pipeline.2. According to the characteristic of piezoelectric acceleration transducer, signal amplification and filtering circuits were designed and LabVIEW virtual instrument technology was used for data collection and signal processing.3. A wavelet transform tools were used for the signal de-noise handling. Analyze noise and leakage signal through db4 wavelet, estimate threshold using long logarithm means, and validate this in wavelet's soft threshold value de-noising method.4. Higher-order statistics (three-order cumulant and bispectrum) were used to analyze leakage signal and accomplish algorithm of leakage detection. Large amount of experiments show the validity with more than 97% accuracy rate.5. Generalized cross correlation (GCC) method were used to the leakage location. Range in 0.05~0.25MPa, experimental results show that absolute errors were less than 1m in 30m location distance.