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.

Uncertainty Management in Lebesgue Sampling-Based Fault Diagnosis and Prognosis

2016 ◽  
Author(s):  
Wuzhao Yan ◽  
Bin Zhang
Keyword(s):  
Fault Diagnosis ◽  
Particle Filtering ◽  
Uncertainty Management ◽  
Remaining Useful Life ◽  
Measurement Noise ◽  
Diagnostic Model ◽  
Contributing Factors ◽  
Diagnosis And Prognosis ◽  
Short Term Prediction ◽  
Fault Diagnosis And Prognosis

This paper develops the uncertainty management of fault diagnosis and prognosis (FDP) in Lebesgue sampling (LS)-based framework with an application to helicopter drivetrain gearbox. In the developed LS-based FDP system, a particle filtering-based FDP algorithm, fault diagnostic model, failure prognostic model, and uncertainty management are discussed. Although uncertainty management has been developed in the traditional Riemann sampling (RS)-based FDP, it needs to be analyzed and managed in a totally different way since the working principle of LS-FDP is fundamentally different from that of RS-FDP. Inaccurate model structure and parameter, measurement noise, process noise, and unknown future loading are major contributing factors of uncertainties in LS-FDP framework. Since the noise in LS-based prognosis is a distribution on time axis while the noise in RS-based prognosis is one on fault state axis, this paper studies the transpose of noise distribution from state domain to time domain. In order to reduce the uncertainty in the prediction of remaining useful life (RUL), model noise and measurement noise terms are adjusted based on a short-term prediction with n steps and correction loop. In this scheme, the priori time distribution at the (t + n)-th Lebesgue state is predicted and stored at the t-th Lebesgue state. Then, at the (t + n)-th Lebesgue state, when the posteriori distribution becomes available, it is compared with the stored priori distribution to manage the uncertainty. The methods for uncertainty management are illustrated by a case study of the prediction of RUL of gearbox. The experimental results show that the uncertainty in the diagnosis and prognosis process of gearbox is properly managed and the confidence interval is decreased, which enhances the confidence level for decision-making and condition-based maintenance.

scholarly journals Special Issue “Advances in Machine Learning and Deep Learning Based Machine Fault Diagnosis and Prognosis”

Processes ◽  
2021 ◽  
Vol 9 (3) ◽  
pp. 532
Author(s):  
Mohand Djeziri ◽  
Marc Bendahan
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Fault Diagnosis ◽  
Special Issue ◽  
Maintenance Strategies ◽  
Corrective Maintenance ◽  
Diagnosis And Prognosis ◽  
Machine Fault Diagnosis ◽  
Failure Prognosis ◽  
Fault Diagnosis And Prognosis

Fault diagnosis and failure prognosis aim to reduce downtime of the systems and to optimise their performance by replacing preventive and corrective maintenance strategies with predictive or conditional ones [...]

scholarly journals Novel Waveforms, Models, Algorithms for Cable Health Monitoring

2019 ◽  
Vol 11 (1) ◽  
Author(s):  
Xuan Wang
Keyword(s):  
Particle Filtering ◽  
Degradation Mechanism ◽  
Accelerated Aging ◽  
Remaining Useful Life ◽  
Cable Insulation ◽  
Diagnosis And Prognosis ◽  
Accelerated Aging Test ◽  
Aging Test ◽  
Insulation Degradation ◽  
Fault Diagnosis And Prognosis

Cables in power generation and delivery are under high thermal stress cycles. Such high temperature can lead to cable insulation degradation, which will reduce the projected lifetime. Existing methods mainly focus on cable fault detection or insulation degradation mechanism. There is no existing tools for diagnosing the insulation degradation level and predicting the remaining useful life of the cable. The goal of my Ph.D. research is to develop reflectometry and data based approaches to monitor the health status of cables. The research will be conducted in three steps: (1) development of reflectometry based method to monitor the cable insulation degradation; (2) feature extraction and cable insulation degradation dynamic modeling based on the accelerated aging test data; (3) development of risksenstive particle filtering based fault diagnosis and prognosis algorithms for cable degradation; and (4) verification and validation the proposed solution with new experiment data and comparison with existing approaches.

scholarly journals Decentralized Fault Diagnosis and Prognosis Scheme for Interconnected Nonlinear Discrete-Time Systems

2020 ◽  
Vol 8 (1) ◽  
Author(s):  
H. Ferdowsi ◽  
S. Jagannathan
Keyword(s):  
Fault Diagnosis ◽  
Discrete Time ◽  
Fault Isolation ◽  
Remaining Useful Life ◽  
Detection Scheme ◽  
Diagnosis And Prognosis ◽  
Discrete Time Systems ◽  
Local Measurements ◽  
Time Systems ◽  
Fault Diagnosis And Prognosis

This paper deals with the design of a decentralized fault diagnosis and prognosis scheme for interconnected nonlinear discrete-time systems which are modelled as the interconnection of several subsystems. For each subsystem, a local fault detector (LFD) is designed based on the dynamic model of the local subsystem and the local states. Each LFD consists of an observer with an online neural network (NN)-based approximator. The online NN approximators only use local measurements as their inputs, and are always turned on and continuously learn the interconnection as well as possible fault function. A fault is detected by comparing the output of each online NN approximator with a predefined threshold instead of using the residual. Derivation of robust detection thresholds and fault detectability conditions are also included. Due to interconnected nature of the overall system, the effect of faults propagate to other subsystems, thus a fault might be detected in more than one subsystem. Upon detection, faults local to the subsystem and from other subsystems are isolated by using a central fault isolation unit which receives detection time information from all LFDs. The proposed scheme also provides the time-to-failure or remaining useful life information by using local measurements. Simulation results provide the effectiveness of the proposed decentralized fault detection scheme.

scholarly journals RUL Estimation Enhancement using Hybrid Deep Learning Methods

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Ikram Remadna ◽  
Labib Sadek Terrissa ◽  
Soheyb Ayad ◽  
Nourddine Zerhouni
Keyword(s):  
Deep Learning ◽  
Short Term Memory ◽  
Hybrid Methods ◽  
Remaining Useful Life ◽  
Health State ◽  
Learning Methods ◽  
Turbofan Engines ◽  
Learning Techniques ◽  
Useful Life ◽  
Behavior Health

The turbofan engine is one of the most critical aircraft components. Its failure may introduce unwanted downtime, expensive repair, and affect safety performance. Therefore, It is essential to accurately detect upcoming failures by predicting the future behavior health state of turbofan engines as well as its Remaining Useful Life. The use of deep learning techniques to estimate Remaining Useful Life has seen a growing interest over the last decade. However, hybrid deep learning methods have not been sufficiently explored yet by researchers.In this paper, we proposed two-hybrid methods combining Convolutional Auto-encoder (CAE), Bi-directional Gated Recurrent Unit (BDGRU), Bi-directional Long-Short Term Memory (BDLSTM), and Convolutional Neural Network (CNN) to enhance the RUL estimation. The results indicate that the hybrid methods exhibit the most reliable RUL prediction accuracy and significantly outperform the most robust predictions in the literature.

Deep learning for fault diagnosis and prognosis in manufacturing systems

2019 ◽  
Vol 110 ◽  
pp. 1-2 ◽  
Author(s):  
Ruqiang Yan ◽  
Xuefeng Chen ◽  
Peng Wang ◽  
Darian M. Onchis
Keyword(s):  
Deep Learning ◽  
Fault Diagnosis ◽  
Manufacturing Systems ◽  
Diagnosis And Prognosis ◽  
Fault Diagnosis And Prognosis

A Particle Filtering Framework for Failure Prognosis

2005 ◽  
Author(s):  
Marcos Orchard ◽  
Biqing Wu ◽  
George Vachtsevanos
Keyword(s):  
Learning Strategies ◽  
Particle Filtering ◽  
Measurement Model ◽  
Remaining Useful Life ◽  
Diagnosis And Prognosis ◽  
Novel Approach ◽  
Posterior Probability Density ◽  
Failure Prognosis ◽  
Fault Diagnosis And Prognosis

Bayesian estimation techniques are finding application domains in machinery fault diagnosis and prognosis of the remaining useful life of a failing component/subsystem. This paper introduces a methodology for accurate and precise prediction of a failing component based on particle filtering and learning strategies. This novel approach employs a state dynamic model and a measurement model to predict the posterior probability density function of the state, i.e., to predict the time evolution of a fault or fatigue damage. It avoids the linearity and Gaussian noise assumption of Kalman filtering and provides a robust framework for long-term prognosis while accounting effectively for uncertainties. Correction terms are estimated in a learning paradigm to improve the accuracy and precision of the algorithm for long-term prediction. The proposed approach is applied to a crack fault and the results support its robustness and superiority.

scholarly journals Remaining Useful Life Estimation through Deep Learning Partial Differential Equation Models: A Framework for Degradation Dynamics Interpretation Using Latent Variables

2021 ◽  
Vol 2021 ◽  
pp. 1-15
Author(s):  
Sergio Cofre-Martel ◽  
Enrique Lopez Droguett ◽  
Mohammad Modarres
Keyword(s):  
Deep Learning ◽  
Latent Variables ◽  
Latent Variable ◽  
Failure Modes ◽  
Remaining Useful Life ◽  
Space Representation ◽  
Health State ◽  
Partial Differential ◽  
Unseen Data ◽  
Useful Life

Remaining useful life (RUL) estimation is one of the main objectives of prognostics and health management (PHM) frameworks. For the past decade, researchers have explored the application of deep learning (DL) regression algorithms to predict the system’s health state behavior based on sensor readings from the monitoring system. Although the state-of-art results have been achieved in benchmark problems, most DL-PHM algorithms are treated as black-box functions, giving little-to-no control over data interpretation. This becomes an issue when the models unknowingly break the governing laws of physics when no constraints are imposed. The latest research efforts have focused on applying complex DL models to achieve low prediction errors rather than studying how they interpret the data’s behavior and the system itself. This paper proposes an open-box approach using a deep neural network framework to explore the physics of a complex system’s degradation through partial differential equations (PDEs). This proposed framework is an attempt to bridge the gap between statistic-based PHM and physics-based PHM. The framework has three stages, and it aims to discover the health state of the system through a latent variable while still providing a RUL estimation. Results show that the latent variable can capture the failure modes of the system. A latent space representation can also be used as a health state estimator through a random forest classifier with up to a 90% performance on new unseen data.

Sign in / Sign up

Export Citation Format

Share Document