Leak detection for galvanized steel pipes due to loosening of screw thread connections based on acoustic emission and neural networks

2017 ◽  
Vol 24 (18) ◽  
pp. 4122-4129 ◽  
Author(s):  
YJ Song ◽  
SZ Li
Keyword(s):  
Neural Networks ◽  
Acoustic Emission ◽  
Engineering Application ◽  
Leak Detection ◽  
Identification Accuracy ◽  
Galvanized Steel ◽  
Screw Thread ◽  
Ann Model ◽  
Pipe System ◽  
Steel Pipes

Galvanized steel pipes with screw thread connections are widely used in indoor gas transportation. In contrast with the failure of pipe tubes, leakage in this system is prone to occur in the screw thread connections. Aiming at this specific engineering application, a method based on acoustic emission (AE) and artificial neural networks (ANNs) is proposed to detect small gas leaks. Experiments are conducted on a specifically designed galvanized steel pipe system with the manipulated leak occurring in the screw thread connection to acquire the raw AE data. The features in the time and frequency domains are extracted and selected to establish an ANN model for leak detection. It has been validated that the developed ANN-based leak detector can achieve an identification accuracy of over 98%. It is also verified that the proposed model is effective even when the AE signals due to a small leak pass over two screw thread connections or an elbow connection.

The Use of Acoustic Emission for Leak Detection in Steel Pipe

2015 ◽  
Vol 771 ◽  
pp. 88-91
Author(s):  
I.B. Ardhana Putra ◽  
Iwan Prasetiyo ◽  
Dewi Permata Sari
Keyword(s):  
Acoustic Emission ◽  
Detection System ◽  
Propagation Constant ◽  
Leak Detection ◽  
Galvanized Steel ◽  
Steel Pipe ◽  
Experimental Result ◽  
Digital Oscilloscope ◽  
Ae Signal ◽  
Experimental Rig

A leak detection system using acoustic emission methods is developed. For this, an experimental rig to detect leak was built using 8” galvanized steel pipe. The length of the pipe is 2 meters. A leak was made with 3 mm diameter and located in 1 meter from the end pipe. The pipe was filled with water and compressed until certain pressure reached. An acoustic emission transducer from Brüel and Kjær type 8313 is mounted on the pipe wall and connected to digital oscilloscope to detect AE signal. The experiment conducted by placing a sensor at a distance of 15 cm, 30 cm, 45 cm, 60 cm, and 75 cm from the position of the leak. Measurements were also performed with the variation of the pressure 3 bar, 4 bars, 5 bars, 6 bars, and 7 bar for those points.Considering acoustic emission wave travelling on pipe is plane wave, leak detection using energy attenuation emission become possible that is different from the method commonly used. Propagation constant is thus required and obtained based on experimental result where the amplitude varies with the spatial and pressure. It is found that for the case considered here. Subsequently, distance of leak location can be determined by the propagation constant and the ratio of energy. Using this method, the error of prediction is about 15.8 %.

Acoustic emission and moving window-improved kernel entropy component analysis for structural condition monitoring of hoisting machinery under various working conditions

2021 ◽  
pp. 147592172110336
Author(s):  
Yang Li ◽  
Feiyun Xu
Keyword(s):  
Acoustic Emission ◽  
Condition Monitoring ◽  
Working Conditions ◽  
Engineering Application ◽  
Component Analysis ◽  
Structural Condition ◽  
Identification Accuracy ◽  
Moving Window ◽  
Feature Information ◽  
Entropy Component

Acoustic emission (AE) has been widely used to the nondestructive evaluation (NDE) and structural health monitoring (SHM) of hoisting machinery recently. Kernel entropy component analysis (KECA) is generally applied to extract the AE features based on its excellent nonlinear ability. However, traditional KECA specifically requires a considerable number of components (e.g. eigenvalues and eigenvectors) to excellently describe the original data, which leads to a reduction in the effect of approximate dimensionality reduction of high-dimensional data, thus causing readily unacceptable condition monitoring result. To overcome this weakness, a novel method named moving window-improved kernel entropy component analysis (MW-IKECA) is proposed in this study for structural condition monitoring of hoisting machinery, which is aimed at extracting more AE feature information and improving the condition identification accuracy. Firstly, a twiddle factor is introduced in the KECA model for the purpose of breaking the restriction that the projection axes originate only from the feature vectors and maximizing the independence between the components. Meanwhile, the moving window local strategy is incorporated into the proposed IKECA to extract more rich and effectiveness AE feature information at different scales. Finally, the Cauchy–Schwarz (CS) statistic is utilized to calculate the similarity between probability density functions and maintain the angular structure of the MW-IKECA feature space for the task of improving the monitoring accuracy and shortening the monitoring time-delay of MW-IKECA. Results of the experimental and practical engineering application validate the effectiveness and superiority of the proposed method in AE-based crane SHM under different working conditions compared with the traditional KECA and some combinatorial methods.

Pretrained Convolutional Neural Networks Perform Well in a Challenging Test Case: Identification of Plant Bugs (Hemiptera: Miridae) Using a Small Number of Training Images

2021 ◽  
Vol 5 (2) ◽  
Author(s):  
Alexander Knyshov ◽  
Samantha Hoang ◽  
Christiane Weirauch
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Training Image ◽  
Species Level ◽  
Identification Accuracy ◽  
Support Vector ◽  
Test Case ◽  
Learning Approaches ◽  
Lower Accuracy ◽  
Neural Network Classifiers

Abstract Automated insect identification systems have been explored for more than two decades but have only recently started to take advantage of powerful and versatile convolutional neural networks (CNNs). While typical CNN applications still require large training image datasets with hundreds of images per taxon, pretrained CNNs recently have been shown to be highly accurate, while being trained on much smaller datasets. We here evaluate the performance of CNN-based machine learning approaches in identifying three curated species-level dorsal habitus datasets for Miridae, the plant bugs. Miridae are of economic importance, but species-level identifications are challenging and typically rely on information other than dorsal habitus (e.g., host plants, locality, genitalic structures). Each dataset contained 2–6 species and 126–246 images in total, with a mean of only 32 images per species for the most difficult dataset. We find that closely related species of plant bugs can be identified with 80–90% accuracy based on their dorsal habitus alone. The pretrained CNN performed 10–20% better than a taxon expert who had access to the same dorsal habitus images. We find that feature extraction protocols (selection and combination of blocks of CNN layers) impact identification accuracy much more than the classifying mechanism (support vector machine and deep neural network classifiers). While our network has much lower accuracy on photographs of live insects (62%), overall results confirm that a pretrained CNN can be straightforwardly adapted to collection-based images for a new taxonomic group and successfully extract relevant features to classify insect species.

scholarly journals Predicting Sooting Propensity of Oxygenated Fuels Using Artificial Neural Networks

Processes ◽  
2021 ◽  
Vol 9 (6) ◽  
pp. 1070
Author(s):  
Abdul Gani Abdul Jameel
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Functional Groups ◽  
Average Error ◽  
Ann Model ◽  
Oh Groups ◽  
Unseen Data ◽  
Learning Capabilities ◽  
Artificial Neural ◽  
Oxygenated Fuels

The self-learning capabilities of artificial neural networks (ANNs) from large datasets have led to their deployment in the prediction of various physical and chemical phenomena. In the present work, an ANN model was developed to predict the yield sooting index (YSI) of oxygenated fuels using the functional group approach. A total of 265 pure compounds comprising six chemical classes, namely paraffins (n and iso), olefins, naphthenes, aromatics, alcohols, and ethers, were dis-assembled into eight constituent functional groups, namely paraffinic CH3 groups, paraffinic CH2 groups, paraffinic CH groups, olefinic –CH=CH2 groups, naphthenic CH-CH2 groups, aromatic C-CH groups, alcoholic OH groups, and ether O groups. These functional groups, in addition to molecular weight and branching index, were used as inputs to develop the ANN model. A neural network with two hidden layers was used to train the model using the Levenberg–Marquardt (ML) training algorithm. The developed model was tested with 15% of the random unseen data points. A regression coefficient (R2) of 0.99 was obtained when the experimental values were compared with the predicted YSI values from the test set. An average error of 3.4% was obtained, which is less than the experimental uncertainty associated with most reported YSI measurements. The developed model can be used for YSI prediction of hydrocarbon fuels containing alcohol and ether-based oxygenates as additives with a high degree of accuracy.

Sign in / Sign up

Export Citation Format

Share Document