A Deep Learning Approach to the Acoustic Condition Monitoring of a Sintering Plant

Pasha, Shahab; Ritz, Christian; Stirling, David; Zulli, Paul; Pinson, David; Chew, Sheng

doi:10.23919/apsipa.2018.8659486

Cited by 9 publications

(6 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Feature calculation was based on the libROSA python package [42]. Other than [13], we did not run machine learning algorithms on the complete data generated via spectral analysis, but rather performed feature extraction and selection 7 to achieve a quicker convergence of the machine learning algorithm. That way, we also obtained means of getting insights into the data by automatic feature selection.…”

Section: Discussionmentioning

confidence: 99%

“…The model was built using the python libraries Keras [45] and TensorFlow [46]. While [13] applied a recurrent neural network (RNN) to their audio data in a similar approach, for a start, we opted for a standard forward one. Comparing several configurations (concerning the number of layers, neurons, and layer types), this network model showed to be sufficiently accurate for our purposes.…”

Section: Discussionmentioning

confidence: 99%

“…There are several examples of prior work facilitating a combination of acoustic condition monitoring and machine learning. Pasha et al present an overview of multiple supervised machine learning techniques (such as SVM, J48, and Deep Learning) used in the scope of acoustic condition monitoring [13]. Acoustic condition monitoring is, for instance, applied to the detection of air leaks in a sintering plant.…”

Section: Machine Learning Approaches For Process Classification In Production Environmentsmentioning

confidence: 99%

“…While such advanced implementations of auditory displays for process monitoring are still restricted, major advances in regard to the incorporation of acoustic information for automated condition monitoring have recently been achieved [9][10][11][12][13]. Research by Fraunhofer Institute for Digital Media Technology in Ilmenau resulted in a service 1 for automated quality assurance combing recording of airborne sounds with machine learning approaches [14].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Auditory augmented process monitoring for cyber physical production systems

Iber

Lechner

Jandl

et al. 2020

Pers Ubiquit Comput

View full text Add to dashboard Cite

We describe two proof-of-concept approaches on the sonification of estimated operation states and conditions focusing on two scenarios: a laboratory setup of a manipulated 3D printer and an industrial setup focusing on the operations of a punching machine. The results of these studies form the basis for the development of an “intelligent” noise protection headphone as part of Cyber Physical Production Systems which provides auditorily augmented information to machine operators and enables radio communication between them. Further application areas are implementations in control rooms (equipped with multi-channel loudspeaker systems) and utilization for training purposes. As a first proof-of-concept, the data stream of error probability estimations regarding partly manipulated 3D printing processes were mapped to three sonification models, providing evidence about momentary operation states. The neural network applied indicates a high accuracy (> 93%) of the error estimation distinguishing between normal and manipulated operation states. None of the manipulated states could be identified by listening. An auditory augmentation, or sonification of these error estimations, provides a considerable benefit to process monitoring. For a second proof-of-concept, setup operations of a punching machine were recorded. Since all operations were apparently flawlessly executed, and there were no errors to be reported, we focused on the identification of operation phases. Each phase of a punching process could be algorithmically distinguished at an estimated probability rate of > 94%. In the auditory display, these phases were represented by different instrumentations of a musical piece in order to allow users to differentiate between operations auditorily.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Machine Learning Approaches For Process Classification In Production Environmentsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Auditory augmented process monitoring for cyber physical production systems

Iber

Lechner

Jandl

et al. 2020

Pers Ubiquit Comput

View full text Add to dashboard Cite

show abstract

“…Jia et al [33] studied a deep neural network (DNN) to directly extract features from the original rolling element bearings and planetary gearboxes data set for fault diagnosis. Pasha et al [34] applied the use of raw acoustic time-frequency spectrograms as input to an RNN for condition onitoring of a sintering plant. Chao et al [35] investigated a 1-D CNN with multi-channel of vibration signals as input for cavitation intensity recognition of high-speed axial piston pumps.…”

Section: Introductionmentioning

confidence: 99%

SMTNet: Hierarchical cavitation intensity recognition based on sub-main transfer network

Yu¹,

Faber²,

Gou³

et al. 2022

Preprint

View full text Add to dashboard Cite

With the rapid development of smart manufacturing, data-driven machinery health management has been of growing attention. As one of the most popular methods in cavitation intensity recognition, deep learning (DL) has been achieving remarkable results. However, current DL methods are restricted by the assumption that all target classes should be treated equally and exclusively. In situations where some classes are more difficult to be distinguished compared to others and where classes might be organised in a hierarchy of categories, current DL methods can not work well. In this study, a novel hierarchical cavitation intensity recognition framework using Sub-Main Transfer Network, termed SMTNet, is proposed to classify acoustic signals of valve cavitation. SMTNet model outputs multiple predictions ordered from coarse to fine along a network corresponding to a hierarchy of target cavitation states. Firstly, a data augmentation method based on Sliding Window with Fast Fourier Transform (Swin-FFT) is developed to solve few-shot problem. Secondly, a 1-D double hierarchical residual block (1-D DHRB) is presented to capture sensitive features of the frequency domain valve acoustic signals. Thirdly, hierarchical multi-label tree is proposed to assist the embedding of the semantic structure of target cavitation states into SMTNet. Fourthly, experience filtering mechanism is proposed to fully learn a prior knowledge of cavitation detection model. Finally, SMTNet has been evaluated on two cavitation datasets without noise (Dataset 1 and Dataset 2), and one cavitation dataset with real noise (Dataset 3) provided by SAMSON AG (Frankfurt). The prediction accurcies of SMTNet for cavitation intensity recognition are as high as 95.32%, 97.16% and 100%, respectively. At the same time, the testing accuracies of SMTNet for cavitation detection are as high as 97.02%, 97.64% and 100%. In addition, SMTNet has also been tested for different frequencies of samples and has achieved excellent results of the highest frequency of samples of mobile phones.

show abstract

Industrial Internet of Things (IIoT) Framework for Real-Time Acoustic Data Analysis

Munirathinam

2021

Algorithms for Intelligent Systems

View full text Add to dashboard Cite

A Deep Learning Approach to the Acoustic Condition Monitoring of a Sintering Plant

Cited by 9 publications

References 10 publications

Auditory augmented process monitoring for cyber physical production systems

Auditory augmented process monitoring for cyber physical production systems

SMTNet: Hierarchical cavitation intensity recognition based on sub-main transfer network

Industrial Internet of Things (IIoT) Framework for Real-Time Acoustic Data Analysis

Contact Info

Product

Resources

About