A Framework for Predicting Remaining Useful Life Curve of Rolling Bearings Under Defect Progression Based on Neural Network and Bayesian Method

Kitai, Masashi; Kobayashi, Takuji; Fujiwara, Hirokazu; Tani, Ryoji; Numao, Masayuki; Fukui, Kenichi

doi:10.1109/access.2021.3073945

Cited by 13 publications

(3 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Zhou et al [16] proposed RUL prediction based on short-time Fourier transform (STFT) and CNN, using CNN to train the fault features extracted by the time-frequency map constructed by STFT, which does not work well for time-frequency analysis of non-smooth signals with large transformations because STFT cannot change the window size during the transformation of the algorithm. KITAI et al [17] proposed RUL prediction based on feature fusion network (FFN) and hierarchical Bayesian regression (HBR), where FFN combines a fully connected network of degenerate index vector (DIV) with CNN. Zhu et al [18] used time-frequency representation (TFR) and multi-scale convolutional neural network (MSCNN) for RUL estimation, and MSCNN can automatically learn to contribute to RUL estimation.…”

Section: Introductionmentioning

confidence: 99%

A novel method for online prediction of the remaining useful life of rolling bearings based on wavelet power spectrogram and Transformer structure

Guo,

Tu,

Zhan

et al. 2023

Eng. Res. Express

View full text Add to dashboard Cite

The vibration signal characteristics of rolling bearings are closely related to the performance decay process, predicting the remaining useful life (RUL) of rolling bearings by vibration signals can effectively prevent the occurrence of bearing failures. In this paper, a deep learning-based method for rolling bearing RUL prediction is proposed. The convolutional neural network (CNN), which is more effective in extracting local information, is combined with the Transformer structure, which is specialized in extracting global information, to deeply explore the complex mapping relationship between signal features in wavelet power spectrogram and bearing RUL. Meanwhile, the method of detecting the first prediction time of rolling bearings based on 3 σ criteria is improved. The proposed method is validated with the XJTU-SY rolling element bearing accelerated life test datasets, as well as compared with other methods to prove its superiority. The results show that the proposed method can effectively extract bearing degradation information and realize the accurate prediction of rolling bearing RUL. The performance-improved rolling bearing RUL prediction model is highly robust and generalizable, which applies to other mechanical parts performance prediction and can be realized for practical applications in industrial fields.

show abstract

Section: Introductionmentioning

confidence: 99%

A novel method for online prediction of the remaining useful life of rolling bearings based on wavelet power spectrogram and Transformer structure

Guo,

Tu,

Zhan

et al. 2023

Eng. Res. Express

View full text Add to dashboard Cite

show abstract

“…The whole degradation data are used to estimate the model parameters. 11 The proportional hazards, 12 proportional intensity 13 exponential, 14 Gaussian process regression, 15 hidden Markov, 16 Bayesian, 17 and Wiener process 18 models have been presented to predict RUL of rolling bearings. In addition, artificial intelligence model automatically constructs input-output relationships without relying on physical mechanisms and empirical knowledge.…”

Section: Introductionmentioning

confidence: 99%

Combining first prediction time identification and time-series feature window for remaining useful life prediction of rolling bearings with limited data

Wang

2023

Proceedings of the Institution of Mechanical Engineers, Part O:

View full text Add to dashboard Cite

Limited data are common in the problem of remaining life prediction (RUL) of rolling bearings, and the distribution of degradation data of rolling bearings under different working conditions is quite different, which makes it difficult to predict the RUL of rolling bearings with limited data. To address this issue, this study combines first prediction time identification (FPT) and time-series feature window (TSFW) for predicting the RUL of rolling bearings with limited data. Firstly, the proper first prediction time is identified by a novel FPT identification method considering root mean square and Kurtosis simultaneously. Subsequently, to accurately capture the sequential characteristics of bearing degradation data, the TSFW is constructed and then adaptively compressed considering degradation factor that is derived mathematically. Based on this, this study employs multi-step ahead rolling prediction strategy with degradation factor from FPT to reveal the future degradation trend and then predict the bearing RUL. Finally, the feasibility and generalization of the proposed method under limited data is validated by carrying out several rolling bearing experiments, and the prediction errors for two representative bearings are 14.46% and 8.06%.

show abstract

“…Attention should also be paid to the development of techniques and algorithms for calculating fatigue wear and the life of spindles based on a set of measured parameters. As for the monitoring of the spindle joint conditions, previous studies [58,82] should be noted; papers [83,84] were devoted to assessing the residual life; however, they cannot be directly applied in online monitoring systems. These issues require separate studies, the results of which may be the subject of future publications.…”

mentioning

confidence: 99%

Telemetry System to Monitor Elastic Torque on Rolling Stand Spindles

Voronin,

Loginov,

Gasiyarova

et al. 2024

JMMP

View full text Add to dashboard Cite

This article outlines the relevance of building online telemetry systems for online monitoring of the technical conditions of rolling mill equipment. Electromechanical systems of the horizontal stand of the plate Mill 5000 are described, when operating in harsh conditions caused by the shock loading when workpieces enter the stand. It is noted that dynamic torque overloads, exceeding the rated motor torque by many-fold, cause the fatigue failure of spindle joints and breakage of rolls. In this regard, the development and implementation of systems for monitoring the elastic torque on spindles are extremely urgent. This issue has long been studied, but the references provide no information on the building principles and hardware composition of such systems. The use of strain gauges connected according to a balanced bridge circuit to measure the elastic torque is justified. This paper’s contribution is the proposed modular principle for building a telemetry monitoring system based on the analysis of known techniques for measuring and transmitting diagnostic data. The developed system structure is provided and the concept of data transfer and processing are explained. This article suggests the inductive power supply of a measuring unit mounted on a shaft without the use of batteries. A hardware structure was developed to be applied in a system for measuring, transmitting, and visualizing signals proportional to the elastic torque, manufactured on the basis of data measuring instruments by leading companies. The specifics of placement and connection of strain gauges are considered. The hardware providing a wireless power supply to the signal encoder and digital data transfer between the transmitter and receiver is described. The results of implementing the system on Mill 5000 are provided. The installation of a telemetry ring and a receiving head for the inductive power supply and data reception is shown. An experimental assessment of the elastic torques occurring when workpieces enter the cage was obtained by implementing a drive control algorithm which provided biting in the drive acceleration mode. The reliability of measuring the elastic torque with an error not exceeding ±5% and the reduction of dynamic loads on the spindle by 1.3–1.5 times due to the elimination of impacts from closing angular gaps in spindle joints was confirmed. This increases the service life of mechanical equipment and reduces the cost of eliminating the accident aftermath. The prospect of modifying the developed system into a cyber-physical system for monitoring the rolling mill’s mechatronic equipment conditions is shown.

show abstract

A Framework for Predicting Remaining Useful Life Curve of Rolling Bearings Under Defect Progression Based on Neural Network and Bayesian Method

Cited by 13 publications

References 24 publications

A novel method for online prediction of the remaining useful life of rolling bearings based on wavelet power spectrogram and Transformer structure

A novel method for online prediction of the remaining useful life of rolling bearings based on wavelet power spectrogram and Transformer structure

Combining first prediction time identification and time-series feature window for remaining useful life prediction of rolling bearings with limited data

Telemetry System to Monitor Elastic Torque on Rolling Stand Spindles

Contact Info

Product

Resources

About