Compressive sensing and sparse decomposition in precision machining process monitoring: From theory to applications

Zhu, Kunpeng; Lin, Xin; Li, Kexuan; Jiang, Lili

doi:10.1016/j.mechatronics.2015.04.017

Cited by 14 publications

(11 citation statements)

References 87 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…time-domain, frequency-domain and time-frequency analysis provide clear physical interpretations, but require high level diagnostic expertise and may fail when incipient or compound faults are developed in machinery operating under varying conditions [1][2]. The latter methods such as artificial neural network, support vector machines and manifold learning [3][4][5] may be more suitable for complex diagnosis problems, but their performance relies strongly on the quality of the hand-crafted features [6][7][8].…”

Section: Introductionmentioning

confidence: 99%

Intelligent Fault Diagnosis for Rotary Machinery Using Transferable Convolutional Neural Network

Chen

Gryllias

2020

IEEE Trans. Ind. Inf.

236

View full text Add to dashboard Cite

Deep neural networks present very competitive results in mechanical fault diagnosis. However, training deep models require high computing power while the performance of deep architectures in extracting discriminative features for decision making often suffers from the lack of sufficient training data. In this paper, a Transferable Convolutional Neural Network (TCNN) is proposed to improve the learning of target tasks. Firstly, a one-dimension CNN is constructed and pre-trained based on large source task datasets. Then a transfer learning strategy is adopted to train a deep model on target tasks by reusing the pre-trained network. Thus, the proposed method not only utilizes the learning power of deep network but also leverages the prior knowledge from the source task. Four case studies are considered and the effects of transfer layers and training sample size on classification effectiveness are investigated. Results show that the proposed method exhibits better performance compared with other algorithms.

show abstract

Section: Introductionmentioning

confidence: 99%

Intelligent Fault Diagnosis for Rotary Machinery Using Transferable Convolutional Neural Network

Chen

Gryllias

2020

IEEE Trans. Ind. Inf.

236

View full text Add to dashboard Cite

show abstract

“…The reconstructed signals preserve the time-frequency representation signatures of the fault signals and the condition states were diagnosed with traditional methods. Zhu [ 15 ] introduced the CS and sparse-decomposition theory for machine and process monitoring. Reference [ 15 ] focused on several state-of-the-art applications, especially, the sparse-decomposition-based fault diagnosis.…”

Section: Introductionmentioning

confidence: 99%

“…Zhu [ 15 ] introduced the CS and sparse-decomposition theory for machine and process monitoring. Reference [ 15 ] focused on several state-of-the-art applications, especially, the sparse-decomposition-based fault diagnosis. The referenced fault-diagnosis methods were all based on the fact that the fault signal could be constructed using weighted linear combinations of the fault samples in the learned dictionary.…”

Section: Introductionmentioning

confidence: 99%

Compressed-Sensing Reconstruction Based on Block Sparse Bayesian Learning in Bearing-Condition Monitoring

Sun

Wen

2017

Sensors

View full text Add to dashboard Cite

Remote monitoring of bearing conditions, using wireless sensor network (WSN), is a developing trend in the industrial field. In complicated industrial environments, WSN face three main constraints: low energy, less memory, and low operational capability. Conventional data-compression methods, which concentrate on data compression only, cannot overcome these limitations. Aiming at these problems, this paper proposed a compressed data acquisition and reconstruction scheme based on Compressed Sensing (CS) which is a novel signal-processing technique and applied it for bearing conditions monitoring via WSN. The compressed data acquisition is realized by projection transformation and can greatly reduce the data volume, which needs the nodes to process and transmit. The reconstruction of original signals is achieved in the host computer by complicated algorithms. The bearing vibration signals not only exhibit the sparsity property, but also have specific structures. This paper introduced the block sparse Bayesian learning (BSBL) algorithm which works by utilizing the block property and inherent structures of signals to reconstruct CS sparsity coefficients of transform domains and further recover the original signals. By using the BSBL, CS reconstruction can be improved remarkably. Experiments and analyses showed that BSBL method has good performance and is suitable for practical bearing-condition monitoring.

show abstract

“…According to Ref. [28], it bridges the gap between the conventional machining and ultraprecision machining. Thus, the recognition of phenomena and process input parameters affecting machined surface quality is of high importance.…”

Section: Introductionmentioning

confidence: 99%

Surface texture formation in precision machining of direct laser deposited tungsten carbide

et al. 2017

View full text Add to dashboard Cite

This paper focuses on an analysis of the surface texture formed during precision machining of tungsten carbide. The work material was fabricated using direct laser deposition (DLD) technology. The experiment included precision milling of tungsten carbide samples with a monolithic torus cubic boron nitride tool and grinding with diamond and alumina cup wheels. An optical surface profiler was applied to the measurements of surface textures and roughness profiles. In addition, the micro-geometry of the milling cutter was measured with the application of an optical device. The surface roughness height was also estimated with the application of a model, which included kinematic-geometric parameters and minimum uncut chip thickness. The research revealed the occurrence of micro-grooves on the machined surface. The surface roughness height calculated on the basis of the traditional kinematic-geometric model was incompatible with the measurements. However, better agreement between the theoretical and experimental values was observed for the minimum uncut chip thickness model.

show abstract

Compressive sensing and sparse decomposition in precision machining process monitoring: From theory to applications

Cited by 14 publications

References 87 publications

Intelligent Fault Diagnosis for Rotary Machinery Using Transferable Convolutional Neural Network

Intelligent Fault Diagnosis for Rotary Machinery Using Transferable Convolutional Neural Network

Compressed-Sensing Reconstruction Based on Block Sparse Bayesian Learning in Bearing-Condition Monitoring

Surface texture formation in precision machining of direct laser deposited tungsten carbide

Contact Info

Product

Resources

About