As detecting the pressure signal is the main method in the real-time leak diagnosis of long pipeline,
an abnormal pressure diagnosis method is proposed to make the leak diagnosis rapidly and accurately. Firstly,
a combination filter algorithm is designed to realize noise reduction. Then, an anomaly detection algorithm is
designed to detect abnormal pressure on the head and tail of the pipeline. Finally, the relevancy of the detected
novelties is computed by Pearson correlation coefficient to identify the leakages. The experimental results show
that the proposed method can rapidly detect the leakage with few false alarms and accurately locate the position
of the leakage.
LIU Jinhai* (刘金海), MA Yanjuan (马艳娟), WU Zhenning (吴振宁), WANG Gang (汪刚)
. Real-Time Pressure Based Diagnosis Method for Oil Pipeline Leakage[J]. Journal of Shanghai Jiaotong University(Science), 2017
, 22(2)
: 233
-239
.
DOI: 10.1007/s12204-017-1826-4
[1] DANETI M. A practical preprocessing treatment forpipeline leak locating improving [C]//Proceedings of2008 IEEE International Conference on Digital ObjectIdentifier. [s.l.]: IEEE, 2008: 9-12.
[2] FERRANTE M, BRUNONE B. Pipe system diagnosisand leak detection by unsteady-state tests. 2.Wavelet analysis [J]. Advances in Water Resources,2003, 26(1): 107-116.
[3] GE C H, YANG H Y, YE H, et al. A fast leak locatingmethod based on wavelet transform [J]. TsinghuaScience & Technology, 2009, 14(5): 551-555.
[4] LI X X, GE M R, DAI X L, et al. Accuracy and reliabilityof multi-GNSS real-time precise positioning:GPS, GLONASS, BEI Dou, and GALIEO [J]. Journalof Geodesy, 2015, 89(6): 607-635.
[5] VALIZADEH S, MOSHIRI B, SALAHSHOOR K.Multiphase pipeline leak detection based on fuzzyclassification [J]. AIP Conference Proceedings, 2009,1159(1): 72-80.
[6] MA Y J, LIU J H, WANG Z G. Modified fuzzymin-max neural network for clustering and its applicationon the pipeline internal inspection data[C]//Proceedings of 35th Chinese Control Conference(CCC). Chengdu: IEEE, 2016: 3509-3513.
[7] ZHANG H G, LIU J H, MA D Z, et al. Data-corebasedfuzzy min-max neural network for pattern classification[J]. IEEE Transactions on Neural Networks,2011, 22(12): 2339-2352.
[8] ZHANG H G,WANG J Y,WANG Z S, et al. Sampleddatasynchronization analysis of Markovian neural networkswith generally incomplete transition rates [J].IEEE Transactions on Neural Networks and LearningSystems, 2017, 28(3): 740-752.
[9] LIU J H, ZHANG H G, FENG J, et al. A new faultdetection and diagnosis method for oil pipeline basedon rough set and neural network [J]. Lecture Notes inComputer Science, 2007, 4493: 561-569.
[10] XU D L, LIU J, YANG J B, et al. Inference and learningmethodology of belief-rule-based expert system forpipeline leak detection [J]. Expert Systems with Applications,2007, 32(1): 103-113.
[11] DA SILVA H V, MOROOKA C K, GUILHERME IR, et al. Leak detection in petroleum pipelines usinga fuzzy system [J]. Journal of Petroleum Science andEngineering, 2005, 49(3/4): 223-238.
[12] ZHANG L B, QIN X Y, WANG Z H, et al. Designinga reliable leak detection system for west productspipeline [J]. Journal of Loss Prevention in the ProcessIndustries, 2009, 22(6): 981-989.
[13] LIU J H, FENG J, TAN L, et al. An algorithm of autoupdatethreshold for singularity analysis of pipelinepressure [J]. Mathematical Problems in Engineering,2013, 2013: 495425.
