Long-distance Oil Pipeline Leakage Detection Key Technologies

The leakage accidents in pipelines transportation not only cause great loss of resoursel, but also pollute the environment, what's more, it may lead to serious casualty. In order to guarantee the pipelines work safely, minimize the losses caused by leakage accidents and improve the level of enterprise's automatic management, it is necessary to study more effective leakage detection and location technology. It is of great significance.Based on the analysis on the pipeline leakage detection and locating technology of negative pressure wave, the paper implements the application of median filtering in denoising on the original pressure data; Through analysis of the characteristics of the pressure data, the paper designs a new classification method to wave segmentation, introduces K-L transform in pattern recognition into the feature extraction of the segmentation of pressure data, and sets ectracted feature as the classification criterion of the wave segmentation, chooses the descending segmentation, and gets the inflexion; Then the paper introduces the multi-resolution analysis and the direct detection algorithm of wave descending inflexion, and integrating with K-L transform, the paper proposes an integral negative pressure wave inflexion detection algorithm—a wave descending inflexion capturing algorithm which integrates direct inflexion detection and K-L transform based on multiresolution.Via the detection to the real data gathered in several pipelines of Daqin Oilfield, the integral algorithm is proved to be of of high accuracy and efficiency and also robust in the capture of waveform descending inflexion in experiment. According to the detected waveform descending inflexion using the integral algorithm, the paper adopts two correlation technology including direct correlation and polar correlation to do correlation analysis tu the negative pressure wave data, and inplements the leakage judgement and location.