A novel pipeline leak detection approach independent of prior failure information. (1st January 2021)

• Record Type: Journal Article
• Authors: Rai, Akhand; Kim, Jong-Myon
• Is Part Of: Measurement. Volume 167(2021)
• Publication Date: 2021-01-01
• Subjects: Pipelines -- Condition monitoring -- Kolmogorov-Smirnov (KS) test -- Gaussian mixture model -- Leak detection; AE acoustic emission -- AI artificial intelligence -- DSTAT d-statistic -- ECDF empirical cumulative distribution function -- EM expectation-minimization -- GMM Gaussian mixture models -- KS Kolmogorov–Smirnov -- LMD local mean decomposition -- ML machine learning -- MLPNN multi-layer perceptron neural network -- MSKS multiscale Kolmogorov–Smirnov -- NLLP natural logarithm of likelihood probability -- PLB pencil lead break tests -- PVC polyvinyl chloride -- RMS root mean square -- SVM support vector machine -- E0(x) ECDF of the sample dataset and corresponding Gaussian distribution -- G0(x) ECDF of the corresponding Gaussian distribution -- H0 null hypothesis for KS test -- H1 alternative hypothesis for KS test -- k number of observations less than or equal to Xi in a given sample dataset -- P (x ≤ Xi) cumulative probability of the observations less than or equal to Xi in a given sample dataset -- dα critical value of KS test d-statistic -- p(x/θ) GMM probability density function -- p(X/θ) likelihood function of GMM model -- g(x/ξm) probability density function for individual GMM component -- λm weight assigned to each Gaussian component -- θ a parameter of the GMM probability density function -- M number of GMM components -- N total number of observations -- X a given sample dataset or time-series -- Yjτ new multi-scale time series at a scale-factor of τ. -- µim mean of the mth Gaussian component -- σim covariance of the mth Gaussian component; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2020.108284
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: A novel pipeline condition monitoring approach without prior leakage information is proposed. Kolmogorov-Smirnov test is utilized to extract features from pipeline acoustic emission signals. Gaussian mixture model is used to approximate the feature space distribution for recognizing the leaks. Abstract: Condition monitoring of pipelines is of importance to detect fluids leakage and associated financial losses and accidents. Artificial intelligence (AI) techniques have been widely used for the pipeline condition assessment. A major limitation of the currently-used supervised AI methods is that they heavily rely on sufficient pipeline failure historical data for their training. To cope with this issue, this paper proposes a health index-oriented approach based on multiscale analysis, Kolmogorov-Smirnov (KS) test, and Gaussian mixture model (GMM) for determining the leakage situation in pipelines. GMM is an unsupervised AI method capable of training itself with pipeline normal condition data. In this study, acoustic emission (AE) signals are first acquired from the pipeline at different pressure conditions. Then, the multiscale analysis and KS test are deployed to extract suitable features from the AE signals. The feature samples corresponding to the pipeline normal condition are used to train the GMM. Finally, the feature samples to be tested are supplied to the GMM and the desired health indicator is obtained. The results confirm the effectiveness of the proposed approach in discriminating the normal and leak conditions as well as the severity of leaks. Further, the GMM classifier trained with features derived from multiscale analysis and the KS test outperforms the GMM trained with crest factor, and mean frequency.