Leakage Detection of Natural Gas Pipelines Based on Neural Networks and Evidence Theory

Making use of the wavelet denoising theory,and through establishing the RBF neural network model,the signal denoised was taken as RBF neural network's input parameter to identify the leakage signal.The D-S evidence theory data fusion was applied to natural gas leakage detection to establish its model;and the leakage signal input was judged at the decision level to locate natural gas pipeline leakage;and the C# language-based leakage detection software employed can quickly and accurately identify these leakages and can locate them.