American Journal of Mathematical and Computational Sciences  
Manuscript Information
 
 
Noise Feature Extraction and Analysis of Silicone Oil Fan on Diesel Engine Based on Non-negative Tensor Factorization
American Journal of Mathematical and Computational Sciences
Vol.5 , No. 1, Publication Date: Feb. 3, 2020, Page: 1-8
927 Views Since February 2, 2020, 309 Downloads Since Feb. 2, 2020
 
 
Authors
 
[1]    

Xiaocong Meng, Engineering Research Institute, Guangxi Yuchai Machinery Co., Ltd, Yulin, China.

[2]    

Xianglin Lu, Engineering Research Institute, Guangxi Yuchai Machinery Co., Ltd, Yulin, China.

[3]    

Haijun Wang, Engineering Research Institute, Guangxi Yuchai Machinery Co., Ltd, Yulin, China.

 
Abstract
 

The rotate speed of the electronic-control silicone oil fan directly affects the passing-by noise of the GB 1495 requirement of the load truck in China. Especially, when the silicone oil fan is directly connected to the main shaft of diesel engine, the noise level is dramatically increases, which makes the spectrum characteristics become complex and difficult to identify for engineer, resulting in the inaccurate control of the rotate speed of the fan. It is necessary to find a high- efficiency method to analyze the frequency of the fault. In this paper, the method of non-negative tensor factorization (NTF) is used to decompose the data for fault analysis. According to the geometric model of NTF, the fixed-point alternating least squares is selected to calculate the decomposed factors after the data transmitted to a three dimensional tensor of the hyperspectral data. While all the factors are computed the sparse local bispectrum can be reconstructed with khatri-rao product between the factors, which also be named basic local feature. Experiments show that the calculation error decreases with the increase of tensor cardinal dimension, the error is still less than 9%, and the maximum proportion of the value of sparse bispectrum extracted is only 0.71%, which proves that the analysis method of non-negative tensor factorization for bispectrum extraction is a well solution.


Keywords
 

Silicone Oil Fan, Non-negative Tensor Factorization, Sparse, Bispectrum Feature


Reference
 
[01]    

GB 1495-2016. Limits and measurement methods of vehicle acceleration noise outside the vehicle.

[02]    

Dal H, Emiro g lu a. O, Bilge H, Et al. Experimental study of the effects of chicken and Turkey biodiesel blends on diesel engine noise emissions [J]. International Journal of Environmental Science and Technology, 2019 (9): 5147-5154.

[03]    

Fei Xiong, Zhiyong Hao, Ruijun Liu, etc. Optimization of NVH performance of diesel engine based on the loudness of radiated noise [J]. Journal of central south university (natural science edition), 2016, 47 (08): 2629-2635.

[04]    

Qingying Ren, LiFeng Wang, Jianqiu Huang, Et al. A Novel Capacitive Temperature Sensor for real-time Monitoring of Temperature in the Silicone Oil Fan Clutch [J]. Key Engineering Materials, 2015, 3928: 662-669.

[05]    

Bangsong Li. Structural design and control analysis of electrically controlled silicone oil fan clutch [J]. Automation applications, 2018 (06): 44-46.

[06]    

Xiu Xu, Zhao Ping, Liyong Zheng. Experimental analysis of an electrically controlled silicon oil clutch fan system on a heavy vehicle [J]. Practical automotive technology, 2017 (12): 217-219.

[07]    

Lin Li, Xiaonan Liu, Xiaolin Feng. Heat Transfer Analysis of Silicone Oil Fan Clutch Based on Finite Element Simulation [J]. Applied Mechanics and Materials, 2015, 3744: 73-76.

[08]    

feng gaoshan. Experimental study on fuel economy improvement of commercial vehicles based on cooling and heat management [J]. Equipment manufacturing technology, 2018 (04: 8-11 +22).

[09]    

Tianpei Feng, Yuedong Sun, Yansong Wang, et al. An evaluation method of noise annoyance degree in non-stationary vehicles based on masking effect [J]. China mechanical engineering, 2017, 28 (24): 2919-2924+2930.

[10]    

Cichocki Andrzej, Zdunek Rafal, Choi Seungjin, et al. Non-negative Tensor Factorization using Alpha and Beta divergences [C]. Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing, Part III. Honolulu, HI: 2007, 1393-1396.

[11]    

H. Arahmane, Mahmoudi A, Hamzaoui, et al. Neutron-gamma discrimination based on support vector machine combined to nonnegative matrix factorization and continuous wavelet transform [J]. Measurement, 2020, 149.

[12]    

Peitao Wang, Zhaoshui He, Kan Xie, et. al. A hybrid algorithm for low-rank approximation of nonnegative matrix factorization [J]. Neurocomputing, 2019, 364: 129-137.

[13]    

Shan Wang. Blind separation of multichannel audio signals based on non-negative tensor decomposition [D]. Southeast university, 2015.

[14]    

M. Laassiri, Hamzaoui, R. Cherkaoui. Nonnegative Tensor Factorization Approach Applied to Fission Chamber's Output Signals Blind Source Separation [J]. Journal of Physics: Conference Series, 2018966 (1): 012063.

[15]    

Vesselinov Velimir V, Alexandrov Boian S, O 'Malley Daniel. The Nonnegative tensor factorization for contaminant source identification. [J]. Journal of contaminant hydrology, 2019220: 66-97.

[16]    

Haijun Wang, Guo Cheng, Guoyong Deng, et al. A Fast Method of Feature Extraction for Lowering Vehicle Pass - By Noise -based on Nonnegative Tucker 3 Decomposition [J]. Journal of Archives of Acoustics, and 2017 (4): 619-629.

[17]    

Yu Zhang, Guoxu Zhou, Qibin Zhao, et al. Fast nonnegative tensor factorization based on accelerated proximal gradient and low-rank approximation [J]. Neurocomputing, 2016, 198: 148-154.





 
  Join Us
 
  Join as Reviewer
 
  Join Editorial Board
 
share:
 
 
Submission
 
 
Membership