A cost effective on-site fault diagnosis method for home appliance rotor failures
- PDF / 1,580,462 Bytes
- 6 Pages / 595.276 x 790.866 pts Page_size
- 83 Downloads / 157 Views
(0123456789().,-volV)(0123456789(). ,- volV)
TECHNICAL PAPER
A cost effective on-site fault diagnosis method for home appliance rotor failures Ji Min Baek1 • Sang Hoon Ji2 • Ja Choon Koo1 Received: 28 October 2019 / Accepted: 18 May 2020 Ó Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Rotating components are one of the most important machine parts used in many industrial applications. Rotating machine commonly used in homes has a washing machine, which occurs with fault frequently by periodic use. Therefore, this study aims to diagnose the washing machine cheaply and accurately by using a smartphone’s microphone. This paper proposes fault diagnosis algorithm developed using FFT, skewness, kurtosis, high pass filter (HPF), A-weighting filter, and support vector machine (SVM). The FFT transforms the time domain into the frequency domain, and skewness and kurtosis analyze unbalance degree of the data. And A-weighting filter is used to filter the data as similar to human hearing and SVM is used to construct diagnostic model. The developed algorithm compensates for the shortcomings of the existing fault diagnosis method and shows high accuracy. In addition, because of using the cheap microphone of the smartphone, it is easy to commercialize due to the low cost, and the accuracy is high enough to show the analysis result almost similar to analysis result of commercial measuring instrument. So, it can be used to diagnose only using the smartphone on the spot.
1 Introduction Fault diagnosis of rotating machinery is one of the areas of keen interest in the industry. In the case of rotating machinery, faults are frequently occurred due to the substandard environment. Especially, the rotating mechanism is a part that is basically included in many industrial applications and home appliances, so the fault of the rotating machinery can have a big impact on the functionality of the entire product (Wang et al. 2018; Glowacz et al. 2018). Therefore, many research are being done in the industry to prevent and diagnose this in advance. The fault diagnosis method for rotating machines have been usually carried out in frequency domain because rotating machines have constant frequency component due to a rotation. Therefore we have analyzed data using FFT that can be seen in the frequency domain (Glowacz 2019; Ciabattoni et al. 2017). Recently, time–frequency techniques such as short-time Fourier transform (STFT) or wavelet transform (WT) have been popularly used & Ja Choon Koo [email protected] 1
School of Mechanical Engineering, Sungkyunkwan University, Suwon 16419, Korea
2
Applied Robot R&D Department, KITECH, Ansan, Korea
(Ciabattoni et al. 2017; Chen et al. 2016; Goumas et al. 2002; Kankar et al. 2011; Zheng et al. 2017; Feng et al. 2017; Chen et al. 2016). However, this method is difficult to accurately discriminate when many noise components are mixed due to the environment. And it takes a long time because it is processed in three dimensions. Another method of fault diagnosis is to use statistical metho
Data Loading...
