Data science applications to improve accuracy of thermocouples

Hu, Zhengbing; Jotsov, Vladimir; Su, Jun; Кочан, Орест; Mykyychuk, Mykola; Кочан, Роман; Sasiuk, Taras

doi:10.1109/is.2016.7737419

Cited by 15 publications

(8 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The authors used the NN classifier [27], [28], [29], LDA [30], [31], and SVM [25], [32], [33]. However, other classifiers could be also used, for example neural network [34], [35].…”

Section: A Msaf-12mentioning

confidence: 99%

Detection of Deterioration of Three-phase Induction Motor using Vibration Signals

Głowacz

Kozik

et al. 2019

Measurement Science Review

106

View full text Add to dashboard Cite

Nowadays detection of deterioration of electrical motors is an important topic of research. Vibration signals often carry diagnostic information of a motor. The authors proposed a setup for the analysis of vibration signals of three-phase induction motors. In this paper rotor fault diagnostic techniques of a three-phase induction motor (TPIM) were presented. The presented techniques used vibration signals and signal processing methods. The authors analyzed the recognition rate of vibration signal readings for 3 states of the TPIM: healthy TPIM, TPIM with 1 broken bar, and TPIM with 2 broken bars. In this paper the authors described a method of the feature extraction of vibration signals Method of Selection of Amplitudes of Frequencies – MSAF-12. Feature vectors were obtained using FFT, MSAF-12, and mean of vector sum. Three methods of classification were used: Nearest Neighbor (NN), Linear Discriminant Analysis (LDA), and Linear Support Vector Machine (LSVM). The obtained results of analyzed classifiers were in the range of 97.61 % – 100 %.

show abstract

“…The authors used the NN classifier [27], [28], [29], LDA [30], [31], and SVM [25], [32], [33]. However, other classifiers could be also used, for example neural network [34], [35].…”

Section: A Msaf-12mentioning

confidence: 99%

Detection of Deterioration of Three-phase Induction Motor using Vibration Signals

Głowacz

Kozik

et al. 2019

Measurement Science Review

106

View full text Add to dashboard Cite

show abstract

“…For the subjective test, a sound localization experiment was performed, and the distance perceived by listeners after applying the estimated HRTF to a sound source was measured. Subsequently, the performance of the proposed method was compared to those of the following other HRTF estimation methods: (1) an HRTF estimation method using average HRTF, referred to as "Average HRTF"; (2) the estimated HRTF by a DNN trained with anthropometric measurements in Section 2 [11], referred to as "DNN(37) HRTF" because there were 37 anthropometric measurements including ear measurements; (3) the estimated HRTF by a DNN trained with only 17 head and torso measurements, referred to as "DNN(17) HRTF." Henceforth, the proposed method is referred to as "CNN-DNN HRTF."…”

Section: Performance Evaluationmentioning

confidence: 99%

“…To overcome these shortcomings, mathematical design methods based on measured HRTFs have also been studied in Reference [7][8][9]. Recently, artificial neural networks have shown meaningful results in various applications, such as temperature estimation and control [10,11], machinery fault diagnosis [12,13], material property prediction [14,15], load forecasting [16], handwritten digit recognition [17], and wind-speed forecasting [18]. In particular, based on biometric information, breast cancer classification [19] and corneal power estimation [20] showed good performance.…”

Section: Introductionmentioning

confidence: 99%

Personalized HRTF Modeling Based on Deep Neural Network Using Anthropometric Measurements and Images of the Ear

Lee

Kim

2018

Applied Sciences

View full text Add to dashboard Cite

This paper proposes a personalized head-related transfer function (HRTF) estimation method based on deep neural networks by using anthropometric measurements and ear images. The proposed method consists of three sub-networks for representing personalized features and estimating the HRTF. As input features for neural networks, the anthropometric measurements regarding the head and torso are used for a feedforward deep neural network (DNN), and the ear images are used for a convolutional neural network (CNN). After that, the outputs of these two sub-networks are merged into another DNN for estimation of the personalized HRTF. To evaluate the performance of the proposed method, objective and subjective evaluations are conducted. For the objective evaluation, the root mean square error (RMSE) and the log spectral distance (LSD) between the reference HRTF and the estimated one are measured. Consequently, the proposed method provides the RMSE of −18.40 dB and LSD of 4.47 dB, which are lower by 0.02 dB and higher by 0.85 dB than the DNN-based method using anthropometric data without pinna measurements, respectively. Next, a sound localization test is performed for the subjective evaluation. As a result, it is shown that the proposed method can localize sound sources with higher accuracy of around 11% and 6% than the average HRTF method and DNN-based method, respectively. In addition, the reductions of the front/back confusion rate by 12.5% and 2.5% are achieved by the proposed method, compared to the average HRTF method and DNN-based method, respectively.

show abstract

“…Technologically, it is very difficult to mitigate the errors of sensors. Therefore, individual calibration [ 10 , 11 ] and artificial intelligence are often used [ 12 , 13 ] to improve the accuracy of sensors [ 14 ]. However, the problem of the instability of sensors cannot always be solved by these means because sensors are exposed to various influences in aggressive environments and acquire large errors [ 15 , 16 , 17 ].…”

Section: Introductionmentioning

confidence: 99%

Energy Structure and Luminescence of CeF3 Crystals

et al. 2021

Self Cite

View full text Add to dashboard Cite

The results of the calculation of the energy band structure and luminescent research of CeF3 crystals are presented. The existence of two 5d1 and 5d2 subbands of the conduction band genetically derived from 5d states of Ce3+ ions with different effective electron masses of 4.9 me and 0.9 me, respectively, is revealed. The large electron effective mass in the 5d1 subband facilitates the localization of electronic excitations forming the 4f-5d cerium Frenkel self-trapped excitons responsible for the CeF3 luminescence. The structure of the excitation spectra of the exciton luminescence peaked at 290 nm, and the defect luminescence at 340 nm confirms the aforementioned calculated features of the conduction band of CeF3 crystals. The peculiarities of the excitation spectra of the luminescence of CaF2:Ce crystals dependent on the cerium concentration are considered with respect to the phase formation possibility of CeF3.

show abstract

Data science applications to improve accuracy of thermocouples

Cited by 15 publications

References 7 publications

Detection of Deterioration of Three-phase Induction Motor using Vibration Signals

Detection of Deterioration of Three-phase Induction Motor using Vibration Signals

Personalized HRTF Modeling Based on Deep Neural Network Using Anthropometric Measurements and Images of the Ear

Energy Structure and Luminescence of CeF3 Crystals

Contact Info

Product

Resources

About