Research On Leakage Detection Method Of Oil Products In Complex Pipe Network

Pipeline transportation, as one of the five biggest transport industries, has decisive position in international oil-gas storage and transportation. But, with the extension of pipe network and lots of pipes being out of repair for long years, leaking accidents occur frequently because of corrosion, disasters and third party damage. To protect our country's property and local environment from leaking oil and gas, studying leak detection technology for pipe network has significance.Traditional single pipe detection method already cannot fulfill the precision and accuracy under this complex pipe network condition. The main factor is the device adjustment of the pipe network disturbed the leaking detection. Then, this thesis aims at the oil products in complex pipe network, presents a pipe network leakage detection method which can eliminate most disturbances of device adjustment.This thesis analyzes all the known device adjustment which can disturb the leaking detection in oil products transportation pipe network, specific to that one equipment operates can cause false alarm, and takes an example to analyze the process that a device adjustment make traditional pipe detection method issue false alarm. To make a leaking detection system can obtain equipment operation information, design and realize an OPC client module using OPC port to collect and save operation information.Present a weighted logical inference model and its design method which can fuse multisensory data to identify conditions. Base on sensors and equipment working conditions in classic monitoring station, build an inference model to identify station's condition. And use specific example, which use this model to process pipe network discretion, to prove this model applies to leaking detection.At last, discuss the specific content implementation of this method in detail, clarify the process of finding special cause in the whole pipe network when monitoring station find unusual situation. Analyze two important parameter which people care most when pipe leak: one is accurate positioning of the pipe leakage point, focus on using negative pressure wave arithmetic to discuss the location of the leakage point when mix oil passing through the pipe and when pipeline with dispersion is shutdown; the other one is the flow of the leak oil at the leakage point, analyze some normal spillage estimate method to design an experience way to estimate leakage, which much more aim at the character of the pipe network.Complex pipe network leakage detection method for the oil products keeps the advantages of wavelet transform method and negative pressure wave method, such as high sensitivity, fast response speed and high location precision. And, it brings in weighted logical inference model which fuses multisensory data to offset their bad accuracy, makes the leakage monitoring system have batter robustness.