A Review of Ball Bearings Fault Size Estimation (FSE), Fault Degradation Estimation (FDE), and Artificial Intelligence Based Approaches during Prognosis

2021 ◽  
Vol 107 ◽  
pp. 3-14
Author(s):  
Henry Hlatshwayo ◽  
Nkosinathi Madushele ◽  
Noor A. Ahmed
Keyword(s):  
High Energy ◽  
Remaining Useful Life ◽  
Maintenance Cost ◽  
Health State ◽  
Ball Bearings ◽  
Size Estimation ◽  
Diagnosis And Prognosis ◽  
Rotary Equipment ◽  
Critical Components ◽  
New Research

Ball bearings are critical components of any industrial rotary equipment. They constitute about 90% of industrial machines’ components – and are thus responsible for the largest proportion of failures – approximately 70-85% of downtime. Defected bearings, while in service, give rise to high vibration amplitudes in rotary equipment, resulting in great reduction in their operational efficiency coupled with high energy consumption. Their premature and inadvertent failure could result in unplanned equipment downtown – thereby causing production loss and increased maintenance cost. Patently, to curtail this, it is vital that their health state is monitored throughout their service life for early faults detection, diagnosis, and prognosis. A knowledge of when a bearing will fail – that is, its remaining useful life (RUL) – can serve as supplement to maintenace decision-making such as determining in advance the time an equipment needs to be taken out-of-service and that can alternatively allow for sufficient lead time for maintenance planning as well. This can correspondingly result in enhancement in rotary systems effectiveness – i.e., availability, reliability, maintainability, and capability. Three popular condition monitoring approaches are signal processing-based approaches namely fault size estimation (FSE) and fault degradation estimation (FDE) as well as artifial intelligent (AI) based approach. It is, however, still a challenge to estimate a bearing fault size and therefore its RUL with high precision based on what has been diagnosed using these approaches. Accordingly, this review holistically explore capabilities and limitations of these approaches from recently published work. The reviewed limations are summarized and serve as new research avenue.

scholarly journals Fault Diagnosis and Prognosis Based on Deep Belief Network and Particle Filtering

2018 ◽  
Vol 10 (1) ◽  
Author(s):  
Guangxing Niu ◽  
Shije Tang ◽  
Zhichao Liu ◽  
Guangquan Zhao ◽  
Bin Zhang
Keyword(s):  
Deep Learning ◽  
Fault Diagnosis ◽  
Particle Filtering ◽  
Learning Algorithms ◽  
Remaining Useful Life ◽  
Health State ◽  
Diagnosis And Prognosis ◽  
Fault Dynamics ◽  
Useful Life ◽  
Fault Diagnosis And Prognosis

Fault diagnosis and prognosis (FDP) plays more and more important role in industries FDP aims to estimate current fault condition and then predict the remaining useful life (RUL). Based on the estimation of health state and RUL, essential decisions on maintenance, control, and planning can be conducted optimally in terms of economy, efficiency, and availability. With the increase of system complexity, it becomes more and more difficult to model the fault dynamics, especially for multiple interacting fault modes and for fault modes that are affected by many internal and external factors. With the development of machine learning and big data, deep learning algorithms become important tools in FDP due to their excellent performance in data processing, information extraction, and automatic modeling. In the past a few years, deep learning algorithms demonstrate outstanding performance in feature extraction and learning fault dynamics. As emerging techniques, their powerful learning capabilities attract more and more attentions and have been extended to various applications. This work presents a novel diagnosis and prognosis methodology which combined deep belief networks (DBNs) and Bayesian estimation. In the proposed work, the DBNs are trained offline using available historical data. The fault dynamic model is then represented by the trained DBNs and modeling uncertainty is described by noise. The integration of DBNs with particle filtering is then developed to provide an estimation of the current fault state and predict the remaining useful life, which is very suitable and efficient for most nonlinear fault models. Experimental studies of lithium-ion batteries are presented to verify the effectiveness of the proposed solution.

A novel methodology for fault size estimation of ball bearings using stator current signal

Measurement ◽  
2021 ◽  
Vol 171 ◽  
pp. 108723
Author(s):  
Chen Wang ◽  
Min Wang ◽  
Bin Yang ◽  
Kaiyu Song ◽  
Yiling Zhang ◽  
...  
Keyword(s):  
Ball Bearings ◽  
Size Estimation ◽  
Current Signal ◽  
Stator Current

scholarly journals Predicting Remaining Useful Life of Rolling Bearings Based on Reliable Degradation Indicator and Temporal Convolution Network with the Quantile Regression

2021 ◽  
Vol 11 (11) ◽  
pp. 4773
Author(s):  
Qiaoping Tian ◽  
Honglei Wang
Keyword(s):  
Quantile Regression ◽  
Probability Density ◽  
Degradation Process ◽  
Remaining Useful Life ◽  
Health State ◽  
Rolling Bearings ◽  
Data Set ◽  
Time Frequency ◽  
Useful Life ◽  
Degradation Indicator

High precision and multi information prediction results of bearing remaining useful life (RUL) can effectively describe the uncertainty of bearing health state and operation state. Aiming at the problem of feature efficient extraction and RUL prediction during rolling bearings operation degradation process, through data reduction and key features mining analysis, a new feature vector based on time-frequency domain joint feature is found to describe the bearings degradation process more comprehensively. In order to keep the effective information without increasing the scale of neural network, a joint feature compression calculation method based on redefined degradation indicator (DI) was proposed to determine the input data set. By combining the temporal convolution network with the quantile regression (TCNQR) algorithm, the probability density forecasting at any time is achieved based on kernel density estimation (KDE) for the conditional distribution of predicted values. The experimental results show that the proposed method can obtain the point prediction results with smaller errors. Compared with the existing quantile regression of long short-term memory network(LSTMQR), the proposed method can construct more accurate prediction interval and probability density curve, which can effectively quantify the uncertainty of bearing running state.

scholarly journals Filling machine preventive maintenance using age replacement method in PT Lucas Djaja

2018 ◽  
Vol 154 ◽  
pp. 01056
Author(s):  
Fifi Herni Mustofa ◽  
Ria Ferdian Utomo ◽  
Kusmaningrum Soemadi
Keyword(s):  
Maintenance Cost ◽  
Time To Failure ◽  
High Failure Rate ◽  
Corrective Maintenance ◽  
Replacement Method ◽  
Critical Components ◽  
Age Replacement ◽  
Engine Maintenance ◽  
A Company ◽  
Mean Time

PT Lucas Djaja is a company engaged in the pharmaceutical industry which produce sterile drugs and non-sterile. Filling machine has a high failure rate and expensive corrective maintenance cost. PT Lucas Djaja has a policy to perform engine maintenance by way of corrective maintenance. The study focused on the critical components, namely bearing R2, bearing 625 and bearing 626. When the replacement of the failure done by the company is currently using the formula mean time to failure with the result of bearing R2 at point 165 days, bearing 625 at a point 205 days, and bearing 626 at a point 182 days. Solutions generated by using age replacement method with minimization of total maintenance cost given on the bearing R2 at a point 60 days, bearing 625 at the point of 80 days and bearing 626 at a point 40 days.

Mooring Integrity Management through Digital Twin and Standardized Inspection Data

2021 ◽  
Author(s):  
Shunsaku Matsumoto ◽  
Vivek Jaiswal ◽  
Tadashi Sugimura ◽  
Shintaro Honjo ◽  
Piotr Szalewski
Keyword(s):  
Simulation Models ◽  
Sensitivity Analyses ◽  
Maintenance Cost ◽  
Digital Twin ◽  
Diagnosis And Prognosis ◽  
Twin System ◽  
Physics Simulation ◽  
Integrity Management ◽  
Risk Visualization ◽  
Inspection Data

Abstract This paper presents a concept of a mooring digital twin frameworkand a standardized inspection datatemplate to enable digital twin. The mooring digital twin framework supports real-time and/or on-demand decision making in mooring integrity management, which minimizes the failure risk while reducing operation and maintenance cost by efficient inspection, monitoring, repair, and strengthening. An industry survey conducted through the DeepStar project 18403 identified a standard template for recording inspection data as a high priority item to enable application of the digital twins for integrity management. Further, mooring chain was selected as a critical mooring component for which a standard inspection template was needed. The characteristics of damage/performance prediction with the proposed mooring digital twin framework are (i) to utilize surrogates and/or reduced-order models trained by high-fidelity physics simulation models, (ii) to combine all available lifecycle data about the mooring system, (iii) to evaluate current and future asset conditions in a systematic way based on the concept of uncertainty quantification (UQ). The general and mooring-specific digital twin development workflows are described with the identified essential data, physics models, and several UQ methodologies such as surrogate modeling, local and global sensitivity analyses, Bayesian prediction etc. Also, the proposed digital twin system architecture is summarized to illustrate the dataflow in digital twin development andutilization. The prototype of mooring digital twin dashboard, web-based risk visualization and advisory system, is developed to demonstrate the capability to visualize the system health diagnosis and prognosis and suggest possible measures/solutions for the high-risk components as a digital twin's insight.

Estimation of Remaining Useful Lifetime of Lithium-Ion Battery Based on Acoustic Emission Measurements

2019 ◽  
Vol 141 (4) ◽  
Author(s):  
Nejra Beganovic ◽  
Dirk Söffker
Keyword(s):  
Acoustic Emission ◽  
Lithium Ion Battery ◽  
Lithium Ion ◽  
High Energy ◽  
High Energy Density ◽  
Storage Device ◽  
Health State ◽  
Battery Management ◽  
Application Fields ◽  
Useful Lifetime

Lithium-ion battery (LIB) utilization as energy storage device in electric and hybrid-electric vehicles, wind turbine systems, a number of portable electrical devices, and in many other application fields is encouraged due to LIB small size alongside high energy density. Monitoring of LIB health state parameters, calculation of additional LIB operating parameters, and the fulfillment of safety requirements are provided through battery management systems. Prediction of remaining useful lifetime (RUL) of LIB and state-of-health (SoH) estimation are identified as still challenging and not completely solved tasks. In this contribution, previous works on RUL/SoH estimation, mainly relied on modeling of underlying electrochemical processes inside LIB, are compared with newly developed approach. The proposed approach utilizes acoustic emission measurements for LIB aging indicators estimation. Developed model for RUL estimation is closely related to frequency spectrum analysis of captured acoustic emission (AE) signal. Features selected from AE measurements are considered as model inputs. The novelty of this approach is the opportunity to estimate RUL/SoH of LIB without necessity to capture some intermediate variables, only indirectly related to RUL/SoH (charging/discharging currents, temperature, and similar). The proposed approach provides the possibility to obtain reliable information about current RUL/SoH without the knowledge about underlying physical processes occurred in LIB. Experimental data sets gathered from LIB aging tests are used for model establishment, training, and validation. The experimental results demonstrate the applicability of the novel approach.

Sign in / Sign up

Export Citation Format

Share Document