Monitoring Performance Degradation Research Articles

Collaborative monitoring of wind turbine performance based on probabilistic power curve comparison

A wind farm is usually equipped with multiple wind turbines of the same type. These wind turbines often work under same complex conditions. Accurate performance degradation monitoring is crucial for ensuring the reliable operation of wind farms and reducing maintenance costs. Motivated by this, this article develops a new wind turbine performance degradation monitoring scheme, which is based on pairwise comparison of the probability power curves of different wind turbines in a wind farm. Firstly, covariate matching is used to eliminate the inherent differences in meteorological variables of different turbines within the same data segment. Next, two probabilistic wind power curves, the quantile power curve and density power curve, model the functional relationship between the meteorological variables and wind power output. Then, deviation vectors are generated by calculating the deviation of probabilistic power curves between each pair of wind turbines. Finally, a directional Hotelling T2 control chart is proposed to monitor the deviation vectors. We apply the new method on the real data of a wind farm in East Britain. Results show that the proposed monitoring technique can monitor wind turbine performance degradation more precisely and comprehensively than the existing approaches.

Remaining useful life prediction for motor systems by iteratively updated wiener process based on closed-loop performance degradation monitoring

Driven by the need for remaining useful life prediction of degraded motor systems with feedback controllers, a real-time updated Wiener stochastic process is adopted to model the performance degradation of motor systems. First, a closed-loop performance index of the motor system is derived incorporating the multiple slow time-varying characteristic of motor parameters. On this basis, the drift coefficient and diffusion coefficient of the Wiener degradation model are updated to obtain the prior maximum likelihood function with available historical data. It is followed by the iterative optimization of nonlinear feature parameter in the Wiener degradation model with taking the prior maximum likelihood function as the cost equation. The effectiveness of proposed remaining useful life prediction architecture for closed-loop motor systems is demonstrated by motor systems.

Physics Informed Machine Learning Approach for Performance Degradation Monitoring of Gas Turbine

The heavy-duty gas turbine is playing an increasingly significant role on power generation due to its lower-emission, higher flexibility and thermo-efficiency. Main subsystems of the gas turbine like compressor, combustor and turbine degrade over the operating time under the harsh environmental conditions, which largely impacts the efficiency and productivity of the system. Therefore, it is critical to develop effective approaches to monitor performance degradation of a heavy-duty gas turbine for system predictive maintenance thus improving the efficiency and productivity of the machine. This paper presents a new physics informed machine learning methodology to predict the degradation of gas turbine by seamlessly integrating thermodynamic heat balancing mechanism, component characteristics, multi-source data and artificial neural network model. The mechanism-based thermodynamic model is established for multiple subsystems considering the balance of flow, mass and energy, and then integrated to a system level for performance simulation of the gas turbine under different conditions. The system model is able to effectively simulate values for those parameters that are not measurable (e.g. GT exhaust flow) or inaccurately measured (e.g. fuel flow). Machine learning based data cleaning approach is employed to preprocess the multivariate raw data of the gas turbine. The difference between design performance data and corrected value obtained from the physics-informed model under ISO conditions is utilized to assess the performance degradation. A Long Short-Term Memory (LSTM) model is established from the fusion of the actual and simulation data to predict the performance degradation of the gas turbine. A comparison study with the classical Nonlinear Autoregressive Network with External Input (NARX) neural network is conducted to demonstrate the advantage of the proposed method. Key Word: Gas Turbine, Thermodynamic Balance, Performance Degradation Predict, Machine Learning, LSTM

Wind turbine performance degradation monitoring using DPGMM and Mahalanobis distance

Real-time monitoring of wind turbine performance degradation can improve the economics and safety of wind farms. Normal operational data can accurately reflect the generation performance of a wind turbine and in the wind-speed coordinate system these normal data constitute the “main power band”. This paper invokes a Dirichlet Process Gaussian Mixture Model (DPGMM) to cluster operational data in each horizontal power bin, and the number of Gaussian components can be determined automatically. The confidence ellipses of Gaussian components can be used to identify the contour of the main power band which is then used as baseline performance model. In the monitoring phase, Mahalanobis distance is used to judge whether new monitoring data lies outside the contour of main power band and thus should be labeled as degraded operational data. When the proportion of such data exceeds a set value in a sliding window, a wind turbine performance degradation alarm is triggered. Degradation degree and rate can quantitatively measure the severity of performance degradation. For an industrial performance degradation case caused by gearbox oil over temperature, the method proposed timely gives alarm only 12 points (2 h) later than the first degraded operational data appears and is proved to be effective.

Failure Detection in Deep Neural Networks for Medical Imaging.

Deep neural networks (DNNs) have started to find their role in the modern healthcare system. DNNs are being developed for diagnosis, prognosis, treatment planning, and outcome prediction for various diseases. With the increasing number of applications of DNNs in modern healthcare, their trustworthiness and reliability are becoming increasingly important. An essential aspect of trustworthiness is detecting the performance degradation and failure of deployed DNNs in medical settings. The softmax output values produced by DNNs are not a calibrated measure of model confidence. Softmax probability numbers are generally higher than the actual model confidence. The model confidence-accuracy gap further increases for wrong predictions and noisy inputs. We employ recently proposed Bayesian deep neural networks (BDNNs) to learn uncertainty in the model parameters. These models simultaneously output the predictions and a measure of confidence in the predictions. By testing these models under various noisy conditions, we show that the (learned) predictive confidence is well calibrated. We use these reliable confidence values for monitoring performance degradation and failure detection in DNNs. We propose two different failure detection methods. In the first method, we define a fixed threshold value based on the behavior of the predictive confidence with changing signal-to-noise ratio (SNR) of the test dataset. The second method learns the threshold value with a neural network. The proposed failure detection mechanisms seamlessly abstain from making decisions when the confidence of the BDNN is below the defined threshold and hold the decision for manual review. Resultantly, the accuracy of the models improves on the unseen test samples. We tested our proposed approach on three medical imaging datasets: PathMNIST, DermaMNIST, and OrganAMNIST, under different levels and types of noise. An increase in the noise of the test images increases the number of abstained samples. BDNNs are inherently robust and show more than 10% accuracy improvement with the proposed failure detection methods. The increased number of abstained samples or an abrupt increase in the predictive variance indicates model performance degradation or possible failure. Our work has the potential to improve the trustworthiness of DNNs and enhance user confidence in the model predictions.

A novel machine learning model for safety risk analysis in flywheel-battery hybrid energy storage system

Flywheel energy storage system (FESS) has been regarded as the most promising hybrid storage technique to manage the battery charging process of electric vehicles. Thanks to properly regulating with the FESS, the battery life can be significantly prolonged. In order to ensure the safety of the hybrid storage system, it is imperative to monitor the mechanical operation condition of the FESS. Because the rolling bearing is a critical mechanical component in the FESS, the performance degradation monitoring and remaining useful life (RUL) prediction of the rolling bearing must be performed. This paper proposes a new machine learning method for the construction of a health indicator to quantitatively evaluate the bearing health status. In this new method, the original feature set is firstly selected through three feature evaluation indicators and the principle component analysis (PCA) is employed to fuse the original feature set as a new health indicator. Then the primary trend of the health indicator is extracted by the empirical mode decomposition (EMD); and the Kriging model-based prediction method is proposed to predict the bearing RUL. The feasibility and superiority of the proposed method is verified through experimental test and analysis result shows that the root mean square error (RMSE) of the prediction is as small as 0.0425.

State degradation evaluation and early fault identification of wind turbine bearings

To solve the shortage of state degradation characteristic and evaluation parameters in the monitoring of the running state of wind turbine bearings, and the low accuracy of early fault identification, a state degradation characteristic index, which combines multi-scale weighted permutation entropy (MWPE) and local linear embedding algorithm (LLE), is proposed for performance degradation monitoring and early fault identification of wind turbine bearings. The degradation index is quantitatively evaluated by four evaluation parameters: the initial time of state degradation, the sensitivity of initial degradation, the failure mutability and the consistency of degradation trend. Through the full cycle acceleration life monitoring data of the bearings, the result show that the proposed MWPE LLE degradation index can effectively characterize the whole process of bearing life cycle from stable operation to degradation to failure. Compared with the quantitative evaluation result of the kernel principal component analysis (KPCA), weighted permutation entropy (WPE), the root mean square (RMS) and kurtosis index, the comprehensive expression ability of MWPE-LLE degradation index is better. It can not only accurately capture the mutation time of state degradation, but also detect the initial degradation time in advance of the four kinds of characteristic indexes, and accurately judge the failure type of the early failure of the test bearing, and the judgment result is consistent with the inspection result after the experiment.

Unified control and detection framework and its applications: A review, some new results, and future perspectives

Initiated three decades ago, integrated design of controllers and fault detectors has continuously attracted research attention. The recent development of the unified control and detection framework with an observer-based residual generator in its core gives a more general form of the previous works. Its applications to residual centred modelling of uncertain control systems, fault detection in feedback control systems with uncertainties, fault-tolerant control (FTC) as well as control performance degradation monitoring, detection and recovery are introduced. In conclusion, some future perspectives are proposed.

Performance Degradation Monitoring and Recovery of Vision-Based Control Systems

As an emerging need in vision-based control that is been intensively deployed in various applications, a systematic way of automatic performance recovery after system maintenance/control performance degradation is highly desired. On the one hand, vision-based control is an interdisciplinary field and the control performance of vision-based control systems is affected by not only uncertain environmental conditions but also the image processing latency, as well as the coordination of control and image processing algorithms. On the other hand, integration of image processing, control, and performance degradation recovery remains an open problem. In light of these observations, this brief deals with performance degradation monitoring and recovery with respect to the aforementioned factors. We first introduce a performance indicator by postfiltering the residual signal with the stable image representation of the feedback controller. It reveals a new interpretation of performance degradation recovery in the input- and output-recovery context. From this perspective, a real-time performance recovery approach is presented to enhance system performance over the lifetime. In the end, a vision-based inverted pendulum control system is implemented to verify the effectiveness of the proposed approach.

Remaining useful life re-prediction methodology based on Wiener process: Subsea Christmas tree system as a case study

With the continuous improvement of the complexity and comprehensive level of the system, its reliability becomes more and more important. The remaining useful life (RUL) estimation method using the degradation model with random effect to describe the degradation process of the system has been widely used such as Wiener process. However, the conventional Wiener-process-based degradation model only considers the current monitoring data but not the historical degradation data, which leads to the inaccuracy of RUL prediction. Furthermore, in engineering, there will always be data missing caused by sensor networks, long life cycle properties of system and so on, leading to unsatisfactory results. This paper contributed a RUL re-prediction method based on Wiener process combining the current monitoring status and historical degradation data of the system. In the initial prediction process, the Wiener process is used to describe the degradation process of the system, the drift coefficient and diffusion coefficient are estimated by Expectation Maximization algorithm (EM algorithm), and the dynamic Bayesian networks (DBNs) model for system performance degradation is established to solve the uncertainty caused by missing data. In the re-prediction process, n groups of performance degradation monitoring data and historical predicted data are combined to calculate the basic degradation in each stage of Wiener process, and the DBNs are used for modeling. The RUL value is obtained by the time difference between the detection point and the predicted fault point, it is determined by the failure threshold finally. A case of subsea Christmas tree system is adopted to demonstrate the proposed approach.

Ultrasonic monitoring performance degradation of lithium ion battery

To monitor the performance and reliability of a lithium-ion battery during charge/discharge cycles, an ultrasonic time-of-flight (ToF) and attenuation measurement was performed. Ultrasonic ToF of the reflected pulse echo increased with the decrease in the state of charge (SoC) and with the increase in number of cycles (i.e., degradation). A health evaluation map relating SoC with ToF at various cycles was constructed. In this map, ultrasonic hysteresis was reported, of which area reflected the magnitude of irreversible electrochemical damage to the battery material. A linear correlation between the ToF and the number of cycles was observed.

Exploring Equipment Electrocardiogram Mechanism for Performance Degradation Monitoring in Smart Manufacturing

Similar to the use of electrocardiogram (ECG) for monitoring heartbeat, this article proposes an equipment electrocardiogram (EECG) mechanism based on fine-grained collection of data during the entire operating duration of the manufacturing equipment, with the purpose of the EECG to reveal the equipment performance degradation in smart manufacturing. First, the system architecture of EECG in smart manufacturing is constructed, and the EECG mechanism is explored, including the granular division of the duration of the production process, the matching strategy for process sequences, and several important working characteristics (e.g., baseline, tolerance, and hotspot). Next, the automatic production line EECG (APL-EECG) is deployed, to optimize the cycle time of the production process and to monitor the performance decay of the equipment online. Finally, the performance of the APL-EECG was validated using a laboratory production line. The experimental results have shown that the APL-EECG can monitor the performance degradation of the equipment in real-time and can improve the production efficiency of the production line. Compared with a previous factory information system, the APL-EECG has shown more accurate and more comprehensive understanding in terms of data for the production process. The EECG mechanism contributes to both equipment fault tracking and optimization of production process. In the long run, APL-EECG can identify potential failures and provide assistance in for preventive maintenance of the equipment.

Design of SH Aging Sensor for Real Time and Application in Sensing Network

One of the most important issues in deep nanoscale regime CMOS circuits is related to the time-dependent performance degradation caused by negative bias temperature instability (NBTI). The integration of online aging sensor is becoming attractive methodologies in monitoring performance degradation of circuit. The sensor can generate a warning signal, early warning of the occurrence of aging faults, to avoid unnecessary losses. To accurately capture the aging fault, a real-time aging sensor is proposed with mirror extraction setup and hold (SH) time method. The proposed aging sensor, which is on the basis of the standard flip-flop (FF), consists of an additional edge detector circuit, a detection window generator circuit, and an output warning circuit. Having the adaptive characteristic of detection window, aging sensor is able to adjust its NBTI effects and improve the stability. Also, the sensing network supports multiple paths online detection from many SH sensors for IP chip applications. Finally, the Camellia IP layout is inserted with 20 aging sensors and is implemented under 65-nm CMOS process. Experimental results demonstrate the effectiveness of area, power, and performance overheads. Compared with other state of the art, hardware efficiency is increased by 46%, and energy is decreased by about 37%.

SeRA: Self-Repairing Architecture for Dark Silicon Era

The lifetime reliability of processors has become a major design constraint in the dark silicon era. Processor reliability issues are mainly due to design defects and aging. Unlike design defects, however, aging faults gradually accumulate over time. Many methods have recently been proposed to monitor the performance degradation of circuits. In this study, an architectural solution that extends the circuit-level age monitoring to processor stages is proposed for monitoring performance degradation. When degradation of a stage quantified as delay of half of the reference clock occurs, a self-repairing mechanism is triggered. This mechanism configures an field programmable gate array (FPGA), which takes over the functions of the degraded unit. The proposed self-repairing mechanism is applied to the stages of the processor data-path. This method (SeRA) has lesser area overhead compared with the state-of-art solutions.

Performance degradation monitoring based on data fusion method for in-service train pneumatic brake system

In order to monitor the brake performance degradation state of the in-service train pneumatic brake system, a data fusion method based on data processing and correlation analysis is proposed in this paper. By using the principal component analysis and analytic hierarchy process to analyze the historical on-board data of one subway line in China, five major indicators of the brake cylinder pressure curve based on seven principal signals are extracted, and the analytic hierarchy model of pneumatic brake system is established. Meanwhile, a standard brake cylinder pressure curve fast fitting tool programmed with MATLAB can realize the rapid curve fitting of historical on-board data under different vehicles, loads, and stations for a domestic subway line, and the algorithm has been validated by the normal and degradation data. By this way, the in-service train pneumatic brake performance degradation state is monitored through the deviation analysis between the standard curve and the fitting one. The method proposed in this paper is also suitable for the analysis of the pneumatic brake system performance of other rail transit trains.

Real-time monitoring energy efficiency and performance degradation of condensing boilers

Condensing boilers achieve higher efficiency than traditional boilers by using waste heat in flue gases to preheat cold return water entering the boiler. Water vapor produced during combustion is condensed into liquid form, thus recovering its latent heat of vaporization, leading to around 10–12% increased efficiency. Many countries have encouraged the use of condensing boilers with financial incentives. It is thus important to develop software tools to assess the correct functioning of the boiler and eventually detect problems. Current monitoring tools are based on boiler static maps and on large sets of historical data, and are unable to assess timely loss of performance due to degradation of the efficiency curve or water leakages. This work develops a set of fault detection and diagnosis tools for dynamic energy efficiency monitoring and assessment in condensing boilers, i.e. performance degradation and faults can be detected using real-time measurements: this real-time feature is particularly relevant because of the limited amount of data that can be stored by state-of-the-art building energy management systems. The monitoring tools are organized as follows: a bimodal parameter estimator to detect deviations of the efficiency of the boiler from nominal values in both condensing and noncondensing mode; a virtual sensor for the estimation of the water mass flow rate; filters to detect actuator and sensor faults, possibly due to control and sensing problems. Most importantly, structural properties for detection and isolation of actuators and sensing faults are given: these properties are crucial to understand which faults can be diagnosed given the available measurements. The effectiveness of these tools is verified via extensive simulations.

Automatic detection of DDoS attacks to notification services

A notification service alerts a large number of recipients to important or emergency events in a timely manner. A denial of service (DoS) attack inserts unnecessary traffic to slow down or chock the notification service and a distributed DoS (DDoS) comes from seemingly various sources. The challenge of automated detection of DDoS attacks lies in distinguishing attacks from temporary surge of normal notification traffic. This paper proposes an 'escalation hierarchy' method to detect such types of DDoS intrusions by monitoring performance degradation at various levels of social events. Our trial of the method on an industrial large-scale notification service showed the effectiveness of our method through automating both consistency checking of measurement data and identification of causes for performance degradation.

Adaptive hidden Markov model-based online learning framework for bearing faulty detection and performance degradation monitoring

This study proposes an adaptive-learning-based method for machine faulty detection and health degradation monitoring. The kernel of the proposed method is an “evolving” model that uses an unsupervised online learning scheme, in which an adaptive hidden Markov model (AHMM) is used for online learning the dynamic health changes of machines in their full life. A statistical index is developed for recognizing the new health states in the machines. Those new health states are then described online by adding of new hidden states in AHMM. Furthermore, the health degradations in machines are quantified online by an AHMM-based health index (HI) that measures the similarity between two density distributions that describe the historic and current health states, respectively. When necessary, the proposed method characterizes the distinct operating modes of the machine and can learn online both abrupt as well as gradual health changes. Our method overcomes some drawbacks of the HIs (e.g., relatively low comprehensibility and applicability) based on fixed monitoring models constructed in the offline phase. Results from its application in a bearing life test reveal that the proposed method is effective in online detection and adaptive assessment of machine health degradation. This study provides a useful guide for developing a condition-based maintenance (CBM) system that uses an online learning method without considerable human intervention.

경계 스캔 기반 온-라인 회로 성능 모니터링 기법

As semiconductor technology has developed, device performance has been improved. However, since device structures became smaller, circuit aging due to operational and environmental conditions can be accelerated. Circuit aging causes a performance degradation and eventually a system error. In reliable systems, a failure due to aging might cause a great disaster. Therefore, these systems need a performance degradation prediction function so that they can take action in advance before a failure occurs. This paper presents an on-line circuit performance degradation monitoring scheme for predicting a failure by detecting performance degradation during circuit normal operation. In our proposed scheme, IEEE 1149.1 output boundary scan cells and TAP controller are reused. The experimental result shows that the proposed architecture can monitor the performance degradation during normal operation without stopping the circuit.

Pilot Tone based CD and PMD Monitoring Technique for Photonic Networks

Objectives: This paper aims at evaluation of pilot tone based Chromatic dispersion and Polarization mode dispersion monitoring technique and to find out a suitable solution for simultaneous monitoring of CD and PMD in dynamic optical network. Method: A pilot tone of 2GHz to 8GHz has been send to each channel of a WDM Link for the monitoring of performance degradation of composite signal during the transmission. CD and PMD causes phase shift of LSB and USB of the amplitude modulated pilot tone which results variation of amplitude of the pilot tone. The amplitude of the pilot tome is the parameter by which measurement of different quantities such as optical power, wavelength, OSNR, CD and PMD can be done in a dynamic optical network. Findings: From the various literatures it has been observed that Optical network is switching from static form to dynamic form. Implementation of plug and play devices, adaptive modulation techniques and various dynamic services introduce complexity in photonic network. Due to highly dynamic and heterogeneous nature of optical network, investigation and implementation of appropriate performance monitoring technique is a major challenge for the researchers. This paper presents a pilot tone based method for simultaneous monitoring of CD and PMD. The proposed method is suitable for dynamic optical network and the simulation results shows low frequency pilot tones provides better resolution for both CD and PMD monitoring technique. The simulation result shows that 2GHz pilot tone provides CD measurement range upto 16000ps/nm and PMD DGD value upto 50ps for the tone power variation of -10 to 0dB. Applications: The proposed techniques can be used for simultaneous CD and PMD monitoring of dynamic WDM optical networks.