A multi-time scale approach to remaining useful life prediction in rolling bearing

Qian, Yuning; Yan, Ruqiang; Gao, Robert X.

doi:10.1016/j.ymssp.2016.06.031

Cited by 188 publications

(76 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, it is necessary to determine a threshold to identify the time of the final failure. Based on the analysis of the references in this field [33,34], the threshold of the final failure ζ is usually selected according to the practice and experience. In details, the value of the health indicator shows a sudden change when a fault occurs, which can be used as the basis for the selection of ζ.…”

Section: Initial and Final Failure Prediction Stagementioning

confidence: 99%

Failure Prediction of the Rotating Machinery Based on CEEMDAN-ApEn Feature and AR-UKF Model

et al. 2020

View full text Add to dashboard Cite

A novel failure prediction method of the rotating machinery is presented in this paper. Firstly, the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) is applied to decompose the vibration signals of the rotating machinery into a number of intrinsic mode functions (IMFs) and a residual (Res), and the metric of maximal information coefficient (MIC) is used to select eligible IMFs to reconstruct signals. Then, the approximate entropy (ApEn)-weighted energy value of the reconstructed signals are calculated to track the degradation process of the rotating machinery. Furthermore, the Chebyshev inequality is introduced to determine the prediction starting time (PST). Finally, the auto regress (AR) model and unscented Kalman filter (UKF) algorithm are used to predict the remaining useful life (RUL) of the rotating machinery. The method is fully evaluated in a test-to-failure experiment. The obtained results show that the proposed method outperforms its counterparts on failure prediction of the rotating machinery.Appl. Sci. 2020, 10, 2056 2 of 18 is confined for the analysis of the stationary signals. Recently, the time-frequency domain analysis method is a hot spot in this field. Empirical mode decomposition (EMD) [7-10], wavelet (WT) [11][12][13][14][15], and neural network (NN) [16][17][18] are common means to extract time-frequency features. Qian et al. used the improved recursive entropy as a health indicator for the performance degradation of bearings [19]. Rai and Upadhyay combined the ensemble empirical mode decomposition (EEMD) with the energy moment entropy to track the degradation process of bearings [20][21][22]. However, the above three types of methods are based on the assumption of linearity. To some extent, vibration signals of the rotating machinery always show a certain degree of non-stationarity and non-linearity because of the constantly changing frictional forces, resistance forces, and loads between components of the rotating machinery.Currently, RUL prediction methods can be categorized into two kinds: model-based and data-driven. The former has to set up a mathematical or physical model which can express the degradation process of the rotating machinery. Liao et al. [23] and Oppenheimer et al.[24] adopted the Paris crack growth model and the Forman crack growth model to predict RUL with the measured data, respectively. Model-based methods can accurately describe the developing trend of damage variables of the rotating machinery according to certain physical laws. However, most mathematical or physical models can't be updated over time, which will result the poor universality. Recently, data-driven methods have become the research focus because they only need the historical data. Therefore, they are quite applicable to the prediction of complex systems for its adjustability. Kalman filter [25,26], Bayesian network [? ], and support vector machine [27] (SVM) are typical data-driven methods used for the RUL prediction of the rotating machinery. However, all above methods a...

show abstract

Section: Initial and Final Failure Prediction Stagementioning

confidence: 99%

Failure Prediction of the Rotating Machinery Based on CEEMDAN-ApEn Feature and AR-UKF Model

et al. 2020

View full text Add to dashboard Cite

show abstract

“…A predictive maintenance (PdM) is a useful tool in the prevention of uninterrupted operation of the machinery . Bearing is a critical element of rotating machines, and its failure leads to catastrophic breakdown of the whole machinery.…”

Section: Introductionmentioning

confidence: 99%

“…A predictive maintenance (PdM) is a useful tool in the prevention of uninterrupted operation of the machinery. 1,2 Bearing is a critical element of rotating machines, and its failure leads to catastrophic breakdown of the whole machinery. Due to this reason, bearing health monitoring such as anomaly detection and remaining useful life (RUL) prediction before the failure occurred have received considerable attention in recent years.…”

Section: Introductionmentioning

confidence: 99%

Hybrid approach for remaining useful life prediction of ball bearings

Wang

Mamo

2019

Quality & Reliability Eng

View full text Add to dashboard Cite

Remaining useful life (RUL) prediction plays an important role in predictive maintenance systems to support decision‐makers for arranging maintenance tasks and related resources. We propose a hybrid approach that is combined an exponential weighted moving average (EWMA) control chart for anomaly detection and machine learning models such as support vector regression (SVR) and random forest regression (RFR) with differential evolution (DE) algorithm to predict the RULs of ball bearings. Here, DE algorithm is used to find the optimal hyperparameters of SVR model. The datasets of ball bearings from the Prognostics Data Repository of NASA are used to compare the prediction performance of different methods. The degradation behavior of training data from the anomaly time to the end of life is used to transfer learning for the testing data in the SVR and RFR models. The results indicate that the proposed methods outperform the other four existing methods in terms of score. Therefore, the proposed hybrid approach is a reliable tool for the RUL prediction of ball bearings.

show abstract

“…Model-based approaches attempt to incorporate physical models of the system into the estimation of the RUL. If the system degradation is modeled precisely, model-based approaches usually exhibit better performance than data-driven approaches [11]. This comes at the expense of having extensive a priori knowledge of the underlying system and having a fine-grained model of the system, which can involve expensive computations.…”

Section: Introductionmentioning

confidence: 99%

A neural network-evolutionary computational framework for remaining useful life estimation of mechanical systems

2019

View full text Add to dashboard Cite

This paper presents a framework for estimating the remaining useful life (RUL) of mechanical systems. The framework consists of a multi-layer perceptron and an evolutionary algorithm for optimizing the data-related parameters. The framework makes use of a strided time window to estimate the RUL for mechanical components. Tuning the data-related parameters can become a very time consuming task. The framework presented here automatically reshapes the data such that the efficiency of the model is increased. Furthermore, the complexity of the model is kept low, e.g. neural networks with few hidden layers and few neurons at each layer. Having simple models has several advantages like short training times and the capacity of being in environments with limited computational resources such as embedded systems. The proposed method is evaluated on the publicly available C-MAPSS dataset [1], its accuracy is compared against other state-of-the art methods for the same dataset.

show abstract

A multi-time scale approach to remaining useful life prediction in rolling bearing

Cited by 188 publications

References 40 publications

Failure Prediction of the Rotating Machinery Based on CEEMDAN-ApEn Feature and AR-UKF Model

Failure Prediction of the Rotating Machinery Based on CEEMDAN-ApEn Feature and AR-UKF Model

Hybrid approach for remaining useful life prediction of ball bearings

A neural network-evolutionary computational framework for remaining useful life estimation of mechanical systems

Contact Info

Product

Resources

About