Pipeline leak diagnosis based on wavelet and statistical features using Dempster–Shafer classifier fusion technique

The key to leak detection and location in water supply pipelines is signal denoising and feature extraction. First, in this paper, an improved spline-local mean decomposition (ISLMD) is proposed to eliminate noise interference. Based on the ISLMD decomposition of a signal, the cross-correlation function between the reference signal and the product functions component can be obtained. And then the PF component containing the leak information can be extracted reasonably. Compared with improved local mean decomposition, the ISLMD has higher accuracy in leak location. Second, an image recognition method using a convolutional neural network for leak detection is proposed, which can better address the problem that the features of different leak apertures or locations are highly similar to each other. The images from the conversion of the reconstructed signals are used as the input of the AlexNet model, which is capable of adaptive extraction of leak signal features. The trained AlexNet model can effectively detect different leak apertures. Finally, the signal time-delay between the upstream and downstream pressure transmitters caused by the leak and propagation of negative pressure wave is determined according to generalized crosscorrelation analysis, and thereby, the leak location is obtained. The experimental results show that the proposed method is effective for leak detection and location. INDEX TERMS Local mean decomposition, convolutional neural network, generalized cross-correlation, leak detection and location, fault detection.

show abstract

“…In our experiment, the pipeline leakage scenarios and pipeline model are established by the use of Flowmaster software [20]. As shown in Fig.…”

Section: Simulation Study a Data Generation By Flowmaster Softwarementioning

confidence: 99%

Leak Detection and Location Based on ISLMD and CNN in a Pipeline

et al. 2019

View full text Add to dashboard Cite

show abstract

“…In our previous studies (Zadkarami et al, 2016, 2017), the input pressure and output flowrate signals were fed into a leak isolation algorithm applied to the Golkhari-Binak oil pipeline. The leakage diagnosis algorithm considered an ANN (artificial neural network) classifier with various feature extraction methods including the statistical techniques, wavelet transform, and a fusion of both methods.…”

Section: Introductionmentioning

confidence: 99%

“…The method was capable of detecting the leakage and identifying its size and location but suffered from the mathematical complexity of the statistical and wavelet feature extraction techniques (Zadkarami et al, 2016). Another study on the Golkhari-Binak oil pipeline was presented by Zadkarami et al (2017), which employed Dempster-Shafer (D-S) classifier fusion to enhance the accuracy of the leakage isolation classifier so that the leakage location and its severity range could be correctly identified. Despite the accurate leakage identification, the results were yield form small datasets.…”

Section: Introductionmentioning

confidence: 99%

Data driven leakage diagnosis for oil pipelines: An integrated approach of factor analysis and deep neural network classifier

Zadkarami

Safavi

Taheri

et al. 2020

Transactions of the Institute of Measurement and Control

Self Cite

View full text Add to dashboard Cite

This paper proposes a novel data-based leakage diagnosis method for big datasets, which identifies the leak occurrence, its size, and its location. Different statistical features are used to express the changes in flow and pressure signals at different leakage scenarios. To improve the performances of the leakage diagnosis approach, factor analysis (FA) is employed for dimension reduction purposes. The optimal features of both pressure and flow signals are then fed as input vectors to a deep neural network (DNN) classifier. The proposed leakage diagnosis method has been applied to the first 20 km of the Golkhari-Binak oil pipeline, located in Iran. The leakage isolation accuracy has been compared with some related works. Simulation results show that the proposed method significantly outperforms others with the average correct classification rate (CCR) of about 98%.

show abstract

“…Among these techniques, fiber optic sensors emerged as the most appropriate for long distance pipelines with no potential safety hazard compared with traditional electrical gauges. From the aspect of algorithms, the fluid transient state due to leakage occurrence is always analyzed, then the leak accidents can be identified, thus locating the leak point, in combination with signal processing techniques and mathematic analysis, including artificial neural network [7,8], support vector machine (SVM) [9,10], harmonic wavelet analysis [11], 2 of 13 etc. Until now, no integrated solution for leakage detection combining both advantages of the fiber sensor and advanced algorithm has been proposed in the literature.…”

Section: Introductionmentioning

confidence: 99%

Pipeline Leak Localization Based on FBG Hoop Strain Sensors Combined with BP Neural Network

Jia

Ren

et al. 2018

Applied Sciences

View full text Add to dashboard Cite

Featured Application: A pipeline leakage localization method using self-developed FBG hoop strain sensors combined with BP neural network is proposed and validated. Abstract: Pipelines function as blood vessels serving to bring life-necessities, so their safe usage is one of the foremost concerns. In our previous work, a fiber Bragg grating (FBG) hoop strain sensor with enhanced sensitivity was developed to measure the pressure drop induced by pipeline leakage. Some hoop strain information during the leakage transient process can be extracted from the amount of FBG hoop strain sensors set along the pipeline. In this paper, an integrated approach of a back-propagation (BP) neural network and hoop strain measurement is first proposed to locate the leak points of the pipeline. Five hoop strain variations are employed as input neurons to achieve pattern recognition so as to predict the leakage point. The RMS error can be as low as 1.01% when choosing appropriate hidden layer neurons. Furthermore, the influence of noise on the network's performance is investigated through superimposing Gaussian noise with a different level. The results demonstrate the feasibility and robustness of the neural network for pipeline leakage localization.

show abstract

Pipeline leak diagnosis based on wavelet and statistical features using Dempster–Shafer classifier fusion technique

Cited by 70 publications

References 11 publications

Leak Detection and Location Based on ISLMD and CNN in a Pipeline

Leak Detection and Location Based on ISLMD and CNN in a Pipeline

Data driven leakage diagnosis for oil pipelines: An integrated approach of factor analysis and deep neural network classifier

Pipeline Leak Localization Based on FBG Hoop Strain Sensors Combined with BP Neural Network

Contact Info

Product

Resources

About