The thesis's intention is to develop a cheap pipeline leaking detection and location system,in this paper, 0-3KHZ Rayleigh Wave is studied.When a pipe is leaking, the signal's average is less than that when the pipe isn't leaking, and the signal's standard deviation is greater. Meanwhile, the signal's power spectrum density is distributed from 1KHZ to 3KHZ and more than one peak-value are found. So, the signal's characteristics both in time and in frequency domains are extracted, and then Artificial Neural Network is used to distinguish a leaking pipeline from a non-leaking one.A new leak location algorithm, which based on three sensors, is advanced in this paper. The new algorithm improves the leak location accuracy.According to previous work, the leak detection and location algorithm is programmed using VC and Matcom. |