Method using L-kurtosis and enhanced clustering-based segmentation to detect faults in axial piston pumps

Gao, Qiang; Xiang, Jiawei; Hou, S.; Tang, Hesheng; Zhong, Yongteng; Ye, Shaogan

doi:10.1016/j.ymssp.2020.107130

Cited by 37 publications

(11 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are many types of beamforming sensor configurations that have been used for source localization purposes, including linear, circular, rectangular, and star configurations, among others. 25,26,27 The more complex configurations typically use more sensors. Among them, the uniform linear sensor array is the simplest form of the configuration for source localization purposes.…”

Section: T-shaped Sensor Cluster-based Source Localizationmentioning

confidence: 99%

A novel T-shaped sensor cluster for acoustic source localization

Gao

Jeon

Park

et al. 2021

Structural Health Monitoring

Self Cite

View full text Add to dashboard Cite

This study proposes a new sensor cluster configuration for localizing an acoustic source in a plate using uniform linear array beamforming and T-shaped sensor clusters. This technique requires neither the properties of the plate material nor a dense array of sensors to find the direction of arrival of the acoustic source. It functions by placing four sensors in a cluster in the shape of the letter “‘T” over a small region of the plate. Uniform linear array beamforming-based source localization is carried out by the constructive interference of different sensor signals. However, this approach has the disadvantage of a very low resolution when the direction of arrival approaches certain values. The L-shaped sensor clusters use the information from the time difference of arrival between the sensors to estimate the direction of arrival, which has a high resolution in all directions except for the direction that is very close to vertical to the cluster. In this study, we numerically and experimentally demonstrate that the proposed T-shaped sensor cluster can accurately localize the acoustic source with no blind area. We also detail its superior performance compared to both uniform linear array beamforming and L-shaped clusters. In the experimental investigation, the maximum deviation of impact source localization was reduced significantly, from 54° to 4° for an aluminum plate, and from 42° to 3° for a composite plate. Furthermore, this novel combined sensor array layout requires only a few sensors, which can significantly reduce the cost of structural health monitoring practice.

show abstract

Section: T-shaped Sensor Cluster-based Source Localizationmentioning

confidence: 99%

A novel T-shaped sensor cluster for acoustic source localization

Gao

Jeon

Park

et al. 2021

Structural Health Monitoring

Self Cite

View full text Add to dashboard Cite

show abstract

“…With a trained AE network, abstract and effective features can be extracted from the raw data to represent useful information [14]. Extracting the depth features contained in the process data can enhance the accuracy and robustness of the monitoring model [15][16][17]. Kong et al proposed a hybrid algorithm of attention recurrent AE for feature extraction, which was successfully applied to rotating machinery diagnosis [18].…”

Section: Introductionmentioning

confidence: 99%

A Fault Identification Method for Electric Submersible Pumps Based on DAE-SVM

Yang

Chen

Zhang

et al. 2022

Shock and Vibration

View full text Add to dashboard Cite

The purpose of this study was to investigate how to detect abnormalities in electric submersible pumps (ESPs) in advance and how to classify the faults by monitoring the production data before pumps break down. Additionally, a new method based on the denoising autoencoder (DAE) and support vector machine (SVM) is proposed. Firstly, the ESP production data were processed and fault-related features were screened using the random forest (RF) algorithm. Secondly, input data were randomly damaged by the addition of noise, a DAE network structure was constructed, and the optimal learning rate, noise reduction coefficient, and other parameters were set. Thirdly, the real-time status of the production data of ESP was monitored with reconstruction errors to detect the point when an abnormality occurs signifying a pending fault. Finally, SVM was used to distinguish the type of fault. Compared with existing fault diagnosis methods, our method not only has the advantages of easy extraction of effective data features, higher accuracy, and strong generalization ability but can also detect an abnormal state indicating a coming fault and identify its type, hence enabling the preparation of an appropriate advance solution.

show abstract

“…Numerous types of research have been carried out for the fault detection of an axial piston pump based on vibration signals. Methods based on cluster analysis [ 13 ], particle swarm optimization [ 14 ], artificial neural networks [ 15 ], hidden semi-Markov [ 16 ], and deep belief networks [ 17 , 18 ] are proposed, and features in the time domain (TD), frequency domain (FD), and time–frequency domains (TFD), including kurtosis, the root mean square, the energy ratio, the coefficients of the wavelet packet transform (WPT), and spectral entropy, are used as model indicators.…”

Section: Introductionmentioning

confidence: 99%

Piston Wear Detection and Feature Selection Based on Vibration Signals Using the Improved Spare Support Vector Machine for Axial Piston Pumps

Xia

Xiang

2022

Materials

Self Cite

View full text Add to dashboard Cite

A piston wear fault is a major failure mode of axial piston pumps, which may decrease their volumetric efficiency and service life. Although fault detection based on machine learning theory can achieve high accuracy, the performance mainly depends on the detection model and feature selection. Feature selection in learning has recently emerged as a crucial issue. Therefore, piston wear detection and feature selection are essential and urgent. In this paper, we propose a vibration signal-based methodology using the improved spare support vector machine, which can integrate the feature selection into the piston wear detection learning process. Forty features are defined to capture the piston wear signature in the time domain, frequency domain, and time–frequency domain. The relevance and impact of sparsity in 40 features are illustrated through the single and multiple statistical feature analysis. Model performance is assessed and the sparse features are discovered. The maximum model testing and training accuracy are 97.50% and 96.60%, respectively. Spare features s10, s12, Ew(8), x7, Ee(5), and Ee(4) are selected and validated. Results show that the proposed methodology is applicable for piston wear detection and feature selection, with high model accuracy and good feature sparsity.

show abstract

Method using L-kurtosis and enhanced clustering-based segmentation to detect faults in axial piston pumps

Cited by 37 publications

References 39 publications

A novel T-shaped sensor cluster for acoustic source localization

A novel T-shaped sensor cluster for acoustic source localization

A Fault Identification Method for Electric Submersible Pumps Based on DAE-SVM

Piston Wear Detection and Feature Selection Based on Vibration Signals Using the Improved Spare Support Vector Machine for Axial Piston Pumps

Contact Info

Product

Resources

About