The Research And Development Of Real-Time Leak-Detection And Alarm System Of Oil Pipeline

This system, in which the advanced small-signal detection and modern signal processing techniques are introduced, has resolved a problem that it is difficult to detect the negative pressure wave caused by the leak of pipeline in the case of low oil pressure at oil stations and successfully detected small negative pressure wave signals in the presence of much industrial noise. And unique signal automatic segmentation and structural pattern recognition techniques are used to judge the leaks, which makes for a decrease in the rate of false alarms and an increase in the detection sensitivity. Based on in-depth analyses of the negative pressure wave structural features, this system brings forward three primary statistic characterization parameters reflecting structural changes of pressure wave signals, and provides a determinant formula for signals within abnormal segments.The advanced wavelet transform technique is introduced into this system to capture the Leak inflexions of the negative pressure wave. And the negative pressure wave signals are multi-resolution analyzed to obtain its low-frequency profile and high-frequency details. Therefore, the said inflexions are stably captured so as to improve the location accuracy. This paper introduces the negative pressure wave detection method and wavelet analyses, one kind of mathematic analysis tools, to process the captured signals. Then the integrative mathematic model is set up for oil leak detection and one new leak detection method is brought forward for oil pipelines, which can collect field data in connection with oil leaks in shorter time so as to achieve the goal of locating the leak points quickly and accurately. Compared with the traditional detection methods, this method is quicker in leak detection, more accurate in location, higher in sensitivity and more reliable.