Showing papers on "Condition monitoring published in 2012"

PRONOSTIA : An experimental platform for bearings accelerated degradation tests.

Abstract: This paper deals with the presentation of an experimental platform called PRONOSTIA, which enables testing, verifying and validating methods related to bearing health assessment, diagnostic and prognostic. The choice of bearings is justified by the fact that most of failures of rotating machines are related to these components. Therefore, bearings can be considered as critical as their failure significantly decreases availability and security of machines. The main objective of PRONOSTIA is to provide real data related to accelerated degradation of bearings performed under constant and/or variable operating conditions, which are online controlled. The operating conditions are characterized by two sensors: a rotating speed sensor and a force sensor. In PRONOSTIA platform, the bearing's health monitoring is ensured by gathering online two types of signals: temperature and vibration (horizontal and vertical accelerometers). Furthermore, the data are recorded with a specific sampling frequency which allows catching all the frequency spectrum of the bearing during its whole degradation. Finally, the monitoring data provided by the sensors can be used for further processing in order to extract relevant features and continuously assess the health condition of the bearing. During the PHM conference, a "IEEE PHM 2012 Prognostic Challenge" is organized. For this purpose, a web link to the degradation data is provided to the competitors to allow them testing and verifying their prognostic methods. The results of each method can then be evaluated regarding its capability to accurately estimate the remaining useful life of the tested bearings.

An artificial neural network method for remaining useful life prediction of equipment subject to condition monitoring

Abstract: Accurate equipment remaining useful life prediction is critical to effective condition based maintenance for improving reliability and reducing overall maintenance cost. In this paper, an artificial neural network (ANN) based method is developed for achieving more accurate remaining useful life prediction of equipment subject to condition monitoring. The ANN model takes the age and multiple condition monitoring measurement values at the present and previous inspection points as the inputs, and the life percentage as the output. A function generalized from the Weibull failure rate function is used to fit each condition monitoring measurement series for a failure history, and the fitted measurement values are used to form the ANN training set so as to reduce the effects of the noise factors that are irrelevant to the equipment degradation. A validation mechanism is introduced in the ANN training process to improve the prediction performance of the ANN model. The proposed ANN method is validated using real-world vibration monitoring data collected from pump bearings in the field. A comparative study is performed between the proposed ANN method and an adapted version of a reported method, and the results demonstrate the advantage of the proposed method in achieving more accurate remaining useful life prediction.

A Data-Driven Failure Prognostics Method Based on Mixture of Gaussians Hidden Markov Models

Abstract: This paper addresses a data-driven prognostics method for the estimation of the Remaining Useful Life (RUL) and the associated confidence value of bearings The proposed method is based on the utilization of the Wavelet Packet Decomposition (WPD) technique, and the Mixture of Gaussians Hidden Markov Models (MoG-HMM) The method relies on two phases: an off-line phase, and an on-line phase During the first phase, the raw data provided by the sensors are first processed to extract features in the form of WPD coefficients The extracted features are then fed to dedicated learning algorithms to estimate the parameters of a corresponding MoG-HMM, which best fits the degradation phenomenon The generated model is exploited during the second phase to continuously assess the current health state of the physical component, and to estimate its RUL value with the associated confidence The developed method is tested on benchmark data taken from the “NASA prognostics data repository” related to several experiments of failures on bearings done under different operating conditions Furthermore, the method is compared to traditional time-feature prognostics and simulation results are given at the end of the paper The results of the developed prognostics method, particularly the estimation of the RUL, can help improving the availability, reliability, and security while reducing the maintenance costs Indeed, the RUL and associated confidence value are relevant information which can be used to take appropriate maintenance and exploitation decisions In practice, this information may help the maintainers to prepare the necessary material and human resources before the occurrence of a failure Thus, the traditional maintenance policies involving corrective and preventive maintenance can be replaced by condition based maintenance

Remaining Useful Life Estimation of Critical Components With Application to Bearings

Abstract: Prognostics activity deals with the estimation of the Remaining Useful Life (RUL) of physical systems based on their current health state and their future operating conditions. RUL estimation can be done by using two main approaches, namely model-based and data-driven approaches. The first approach is based on the utilization of physics of failure models of the degradation, while the second approach is based on the transformation of the data provided by the sensors into models that represent the behavior of the degradation. This paper deals with a data-driven prognostics method, where the RUL of the physical system is assessed depending on its critical component. Once the critical component is identified, and the appropriate sensors installed, the data provided by these sensors are exploited to model the degradation's behavior. For this purpose, Mixture of Gaussians Hidden Markov Models (MoG-HMMs), represented by Dynamic Bayesian Networks (DBNs), are used as a modeling tool. MoG-HMMs allow us to represent the evolution of the component's health condition by hidden states by using temporal or frequency features extracted from the raw signals provided by the sensors. The prognostics process is then done in two phases: a learning phase to generate the behavior model, and an exploitation phase to estimate the current health state and calculate the RUL. Furthermore, the performance of the proposed method is verified by implementing prognostics performance metrics, such as accuracy, precision, and prediction horizon. Finally, the proposed method is applied to real data corresponding to the accelerated life of bearings, and experimental results are discussed.

Novel Industrial Wireless Sensor Networks for Machine Condition Monitoring and Fault Diagnosis

Abstract: This paper proposes a novel industrial wireless sensor network (IWSN) for industrial machine condition monitoring and fault diagnosis. In this paper, the induction motor is taken as an example of monitored industrial equipment due to its wide use in industrial processes. Motor stator current and vibration signals are measured for further processing and analysis. On-sensor node feature extraction and on-sensor fault diagnosis using neural networks are then investigated to address the tension between the higher system requirements of IWSNs and the resource-constrained characteristics of sensor nodes. A two-step classifier fusion approach using Dempster-Shafer theory is also explored to increase diagnosis result quality. Four motor operating conditions-normal without load, normal with load, loose feet, and mass imbalance-are monitored to evaluate the proposed system. Experimental results show that, compared with raw data transmission, on-sensor fault diagnosis could reduce payload transmission data by 99%, decrease node energy consumption by 97%, and prolong node lifetime from 106 to 150 h, an increase of 43%. The final fault diagnosis results using the proposed classifier fusion approach give a result certainty of at least 97.5%. To leverage the advantages of on-sensor fault diagnosis, another system operating mode is explored, which only transmits the fault diagnosis result when a fault happens or at a fixed interval. For this mode, the node lifetime reaches 73 days if sensor nodes transmit diagnosis results once per hour.

Rotational Machine Health Monitoring and Fault Detection Using EMD-Based Acoustic Emission Feature Quantification

Abstract: Acoustic emission (AE)-signal-based techniques have recently been attracting researchers' attention to rotational machine health monitoring and diagnostics due to the advantages of the AE signals over the extensively used vibration signals. Unlike vibration-based methods, the AE-based techniques are in their infant stage of development. From the perspective of machine health monitoring and fault detection, developing an AE-based methodology is important. In this paper, a methodology for rotational machine health monitoring and fault detection using empirical mode decomposition (EMD)-based AE feature quantification is presented. The methodology incorporates a threshold-based denoising technique into EMD to increase the signal-to-noise ratio of the AE bursts. Multiple features are extracted from the denoised signals and then fused into a single compressed AE feature. The compressed AE features are then used for fault detection based on a statistical method. A gear fault detection case study is conducted on a notional split-torque gearbox using AE signals to demonstrate the effectiveness of the methodology. A fault detection performance comparison using the compressed AE features with the existing EMD-based AE features reported in the literature is also conducted.

A Novel Online Technique to Detect Power Transformer Winding Faults

Abstract: Frequency-response analysis (FRA) has been growing in popularity in recent times as a tool to detect mechanical deformation within power transformers. To conduct the test, the transformer has to be taken out of service which may cause interruption to the electricity grid. Moreover, because FRA relies on graphical analysis, it calls for an expert to analyze the results. As so far, there is no standard code for FRA interpretation worldwide. In this paper, a novel online technique is introduced to detect the internal faults within a power transformer by constructing the voltage-current (V - I) locus diagram to provide a current state of the transformer. The technique does not call for any special equipment as it uses the existing metering devices attached to any power transformer to monitor the input voltage, output voltage, and the input current at the power frequency and, hence, online monitoring can be realized. Various types of faults have been simulated to assess its impact on the proposed locus. A Matlab code based on digital image processing is developed to calculate any deviation of the V - I locus with respect to the reference one and to identify the type of fault. The proposed technique is easy to be implemented and automated so that the requirement for expert personnel can be eliminated.

Wind Turbine Generator Condition-Monitoring Using Temperature Trend Analysis

Abstract: Condition monitoring can greatly reduce the maintenance cost for a wind turbine. In this paper, a new condition-monitoring method based on the nonlinear state estimate technique for a wind turbine generator is proposed. The technique is used to construct the normal behavior model of the electrical generator temperature. A new and improved memory matrix construction method is adopted to achieve better coverage of the generator's normal operational space. Generator incipient failure is indicated when the residuals between model estimates and the measured generator temperature become significant. Moving window averaging is used to detect statistically significant changes of the residual mean value and standard deviation in an effective manner; when these parameters exceed predefined thresholds, an incipient failure is flagged. Examples based on data from the Supervisory Control and Data Acquisition system at a wind farm located at Zhangjiakou in northern China have been used to validate the approach and examine its sensitivity to key factors that influence the performance of the approach. It is demonstrated that the technique can identify dangerous generator over temperature before damage has occurred that results in complete shutdown of the turbine.

Fault Analysis and Condition Monitoring of the Wind Turbine Gearbox

Abstract: Data mining algorithms and statistical methods are applied to analyze the jerk data obtained from monitoring the gearbox of a wind turbine. Two types of analyses are performed-failure component identification and monitoring vibration excitement. In failure component identification, the failed stages of the gearbox are identified in time-domain analysis and frequency-domain analysis. In the time domain, correlation coefficient and clustering analysis are applied. The fast Fourier transformation with time windows is utilized to analyze the frequency data. To monitor the vibration excitement of the gearbox in its high-speed stage, data mining algorithms and statistical quality control theory are combined to develop a monitoring model. The capability of the monitoring model to detect changes in the gearbox vibration excitement is validated by the collected data.

Condition Monitoring Power Module Solder Fatigue Using Inverter Harmonic Identification

Abstract: Condition monitoring power semiconductor devices can inform converter maintenance and reduce damage. This paper presents a method to monitor solder fatigue in a voltage source inverter insulated gate bipolar transistor power module by detecting the change of an inverter output harmonic. It is shown that low-order harmonics, caused by nonideal switching, are affected by the device junction temperature, which in turn depends upon module solder condition. To improve the detection accuracy of the phenomenon, the inverter controller is set to cause harmonic resonance at the target harmonic frequency. The would-be resonance is suppressed by an outer control loop where the control action can be used as the condition monitoring signal. Simulation and experiment are presented to validate the method and evaluate its performance in operation.

A resilient condition assessment monitoring system

Abstract: An architecture and supporting methods are presented for the implementation of a resilient condition assessment monitoring system that can adaptively accommodate both cyber and physical anomalies to a monitored system under observation. In particular, the architecture includes three layers: information, assessment, and sensor selection. The information layer estimates probability distributions of process variables based on sensor measurements and assessments of the quality of sensor data. Based on these estimates, the assessment layer then employs probabilistic reasoning methods to assess the plant health. The sensor selection layer selects sensors so that assessments of the plant condition can be made within desired time periods. Resilient features of the developed system are then illustrated by simulations of a simplified power plant model, where a large portion of the sensors are under attack.

Health Condition Monitoring of Machines Based on Hidden Markov Model and Contribution Analysis

Abstract: Degradation parameter from normal to failure condition of machine part or system is needed as an object of health monitoring in condition-based maintenance (CBM). This paper proposes a hidden Markov model (HMM) and contribution-analysis-based method to assess the machine health degradation. A dynamic principal component analysis (DPCA) is used to extract effective features from vibration signals, where inherent signal autocorrelation is considered. A novel machine health assessment indication, HMM-based Mahalanobis distance is proposed to provide a comprehensible indication for quantifying machine health states. A variable-replacing-based contribution analysis method is developed to discover the effective features that are responsible for the detection and assessment of machine health degradation in its whole life. The experimental results based on a bearing test bed show the plausibility and effectiveness of the proposed methods, which can be considered as the machine health degradation monitoring model.

Prediction of Machine Health Condition Using Neuro-Fuzzy and Bayesian Algorithms

Abstract: This paper proposes a novel approach for machine health condition prognosis based on neuro-fuzzy systems (NFSs) and Bayesian algorithms. The NFS, after training with machine condition data, is employed as a prognostic model to forecast the evolution of the machine fault state with time. An online model update scheme is developed on the basis of the probability density function (PDF) of the NFS residuals between the actual and predicted condition data. Bayesian estimation algorithms adopt the model's predicted data as prior information in combination with online measurements to update the degree of belief in the forecasting estimations. In order to simplify the implementation of the proposed approach, a recursive Bayesian algorithm called particle filtering is utilized to calculate in real time a posterior PDF by a set of random samples (or particles) with associated weights. When new data become available, the weights of all particles are updated, and then, predictions are carried out, which form the PDF of the predicted estimations. The developed method is evaluated via two experimental cases-a cracked carrier plate and a faulty bearing. The prediction performance is compared with three prevalent machine condition predictors-recurrent neural networks, NFSs, and recurrent NFSs. The results demonstrate that the proposed approach can predict machine conditions more accurately.

Diagnosis of Three-Phase Electrical Machines Using Multidimensional Demodulation Techniques

Abstract: This paper deals with the diagnosis of three-phase electrical machines and focuses on failures that lead to stator-current modulation. To detect a failure, we propose a new method based on stator-current demodulation. By exploiting the configuration of three-phase machines, we demonstrate that the demodulation can be efficiently performed with low-complexity multidimensional transforms such as the Concordia transform (CT) or the principal component analysis (PCA). From a practical point of view, we also prove that PCA-based demodulation is more attractive than CT. After demodulation, we propose two statistical criteria aiming at measuring the failure severity from the demodulated signals. Simulations and experimental results highlight the good performance of the proposed approach for condition monitoring.

A Data-Mining Approach to Monitoring Wind Turbines

Abstract: The rapid expansion of wind farms has generated interest in operations and maintenance. An operating wind turbine undergoes various state changes, including transformation from a normal to a fault mode. Condition-based maintenance tools are needed to identify potential faults in the system. The prediction of turbine fault modes is of particular interest. In this research, data-mining algorithms are employed to construct prediction models for wind turbine faults. A three-stage prediction process is followed: 1) prediction of a fault of any kind; 2) prediction of specific faults of the system; and 3) identification on unseen faults. A comparative analysis of various data-mining algorithms is reported based on the data collected at a large wind farm. Random forest algorithm models provided the best accuracy among all algorithms tested. The robustness of the predictive model is validated for faults that have occurred at turbines with previously unseen data. The research results discussed in this paper have been derived from data collected at 17 wind turbines.

Wind Turbine Gearbox Condition Monitoring Round Robin Study - Vibration Analysis

Abstract: The Gearbox Reliability Collaborative (GRC) at the National Wind Technology Center (NWTC) tested two identical gearboxes. One was tested on the NWTCs 2.5 MW dynamometer and the other was field tested in a turbine in a nearby wind plant. In the field, the test gearbox experienced two oil loss events that resulted in damage to its internal bearings and gears. Since the damage was not severe, the test gearbox was removed from the field and retested in the NWTCs dynamometer before it was disassembled. During the dynamometer retest, some vibration data along with testing condition information were collected. These data enabled NREL to launch a Wind Turbine Gearbox Condition Monitoring Round Robin project, as described in this report. The main objective of this project was to evaluate different vibration analysis algorithms used in wind turbine condition monitoring (CM) and find out whether the typical practices are effective. With involvement of both academic researchers and industrial partners, the project sets an example on providing cutting edge research results back to industry.

Condition monitoring of wind turbine induction generators with rotor electrical asymmetry

Abstract: This study investigates the condition monitoring of wind turbine wound rotor and doubly fed induction generators with rotor electrical asymmetries by analysis of stator current and total power spectra. The research is verified using experimental data measured on two different test rigs and numerical predictions obtained from a time-stepping electromagnetic model. A steady-state study of current and power spectra for healthy and faulty conditions is performed to identify fault-specific signal changes and consistent slip-dependent fault-indicators on both test rigs. To enable real-time fault frequency tracking, a set of concise analytical expressions, describing fault frequency variation with operating speed, were defined and validated by measurement. A variable speed study, representative of real wind turbine operations, of current and power frequency components for healthy and faulty conditions was then carried out on one test rig, which could simulate wind conditions. The current and power fault frequency tracking previously identified achieved reliable fault detection for two realistic wind turbine generator fault scenarios of differing severity. Conclusions are drawn on the relative merits of current and power signal analysis when used for wind turbine wound rotor induction machine fault detection and diagnosis.

Wireless Measurement System for Structural Health Monitoring With High Time-Synchronization Accuracy

Abstract: Structural health monitoring (SHM) systems have excellent potential to improve the regular operation and maintenance of structures. Wireless networks (WNs) have been used to avoid the high cost of traditional generic wired systems. The most important limitation of SHM wireless systems is time-synchronization accuracy, scalability, and reliability. A complete wireless system for structural identification under environmental load is designed, implemented, deployed, and tested on three different real bridges. Our contribution ranges from the hardware to the graphical front end. System goal is to avoid the main limitations of WNs for SHM particularly in regard to reliability, scalability, and synchronization. We reduce spatial jitter to 125 ns, far below the 120 μs required for high-precision acquisition systems and much better than the 10-μs current solutions, without adding complexity. The system is scalable to a large number of nodes to allow for dense sensor coverage of real-world structures, only limited by a compromise between measurement length and mandatory time to obtain the final result. The system addresses a myriad of problems encountered in a real deployment under difficult conditions, rather than a simulation or laboratory test bed.

A Physically Segmented Hidden Markov Model Approach for Continuous Tool Condition Monitoring: Diagnostics and Prognostics

Abstract: In this paper, a temporal probabilistic approach based on the hidden Markov model (HMM), named physically segmented HMM with continuous output, is introduced for continuous tool condition monitoring in machinery systems. The proposed approach has the advantage of providing an explicit relationship between the actual health states and the hidden state values. The provided relationship is further exploited for formulation and parameter estimation in the proposed approach. The introduced approach is tested for continuous tool wear prediction in a computer numerical control milling machine and compared with two well-established neural network (NN) approaches, namely, multilayer perceptron and Elman network. In the experimental study, the prediction results are provided and compared after adopting appropriate hyper-parameter values for all the approaches by cross-validation. Based on the experimental results, physically segmented HMM approach outperforms the NN approaches. Moreover, the prognosis ability of the proposed approach is studied.

Modern techniques for condition monitoring of railway vehicle dynamics

Abstract: A modern railway system relies on sophisticated monitoring systems for maintenance and renewal activities. Some of the existing conditions monitoring techniques perform fault detection using advanced filtering, system identification and signal analysis methods. These theoretical approaches do not require complex mathematical models of the system and can overcome potential difficulties associated with nonlinearities and parameter variations in the system. Practical applications of condition monitoring tools use sensors which are mounted either on the track or rolling stock. For instance, monitoring wheelset dynamics could be done through the use of track-mounted sensors, while vehicle-based sensors are preferred for monitoring the train infrastructure. This paper attempts to collate and critically appraise the modern techniques used for condition monitoring of railway vehicle dynamics by analysing the advantages and shortcomings of these methods.