期刊名称：International Journal of Signal Processing, Image Processing and Pattern Recognition
印刷版ISSN：2005-4254
出版年度：2013
卷号：6
期号：6
页码：129-140
DOI：10.14257/ijsip.2013.6.6.13
出版社：SERSC
摘要：For natural gas pipeline, it has a leak or not is critical. The most commonly problems in the pipeline leak detection methods are the difficulties to identify, inaccuracy to locate, thus, the natural gas pipeline detection is difficult to be applied, therefore, the use of neural network multi-sensor data fusion of the natural gas pipeline leak detection is particularly important. In this paper, the method is proposed based on RBF neural network and the data fusion of D-S evidence theory for detecting the pipeline leak. Extracting neural network's input parameters through wavelet denoising, then substitute them into neural network and calculate them by multi-sensor data fusion algorithm so as to acquire leaking information
关键词：Leak detection; RBF neural network; Wavelet denoising; Data fusion; D-S ; evidence theory
