Search

BackgroundTo classify chest computed tomography (CT) images as positive or negative for coronavirus disease 2019 (COVID-19) quickly and accurately, researchers attempted to develop effective models by using medical images.ResultsA convolutional neural network (CNN) ensemble model was developed for classifying chest CT images as positive or negative for COVID-19. To classify chest CT images acquired from COVID-19 patients, the proposed COVID19-CNN ensemble model combines the use of multiple trained CNN models with a majority voting strategy. The CNN models were trained to classify chest CT images by transfer learning from well-known pre-trained CNN models and by applying their algorithm hyperparameters as appropriate. The combination of algorithm hyperparameters for a pre-trained CNN model was determined by uniform experimental design. The chest CT images (405 from COVID-19 patients and 397 from healthy patients) used for training and performance testing of the COVID19-CNN ensemble model were obtained from an earlier study by Hu in 2020. Experiments showed that, the COVID19-CNN ensemble model achieved 96.7% accuracy in classifying CT images as COVID-19 positive or negative, which was superior to the accuracies obtained by the individual trained CNN models. Other performance measures (i.e., precision, recall, specificity, and F1-score) obtained bythe COVID19-CNN ensemble model were higher than those obtained by individual trained CNN models.ConclusionsThe COVID19-CNN ensemble model had superior accuracy and excellent capability in classifying chest CT images as COVID-19 positive or negative.

For the data-driven remaining useful life (RUL) prediction for rolling bearings, the traditional machine learning-based methods generally provide insufficient feature representation and adaptive extraction. Although deep learning-based RUL prediction methods can solve these problems to some extent, they still do not yield satisfactory predictive results due to less degradation data and inconsistent data distribution among different bearings. To solve these problems, a new RUL prediction method based on deep feature representation and transfer learning is proposed in this paper. This method includes an off-line stage and an online stage. In the off-line stage, the Hilbert–Huang transform marginal spectra of the raw vibration signal of auxiliary bearings are first calculated as the input, and then contractive denoising autoencoder is introduced to extract deep features with good and stable fault representation. Second, by using the obtained deep features and Pearson’s correlation coefficient, a new health condition assessment method is proposed to divide the whole life of each bearing into a normal state and a fast-degradation state. Finally, using the extracted deep features and their RUL values, an RUL prediction model for the fast-degradation state is trained by means of a least-square support vector machine. In the online stage, a kind of transfer learning algorithm, i.e., transfer component analysis, is introduced to sequentially adjust the features of target bearing from auxiliary bearings, and then the corresponding RUL is predicted using the corrected features. Results using the PHM Challenging 2012 data set show a significant performance improvement when using the proposed method in terms of predictive accuracy and numerical stability.

In a big data environment, the data from long-term monitoring are of great importance in intelligent diagnoses, prognostic and health management of industrial components. However, in practice, it is difficult to ensure high reconstructive accuracy making the preservation performance of low-frequency information (e.g., periodic fault impulses) stable. To address this issue, this article proposes a new data reconstruction method that combines a smoothing sparse low-rank matrix (SSLRM) with online dictionary learning. This article first presents the design of SSLRM model for decomposing the raw data into a low resonance component (LRC) and a high resonance component (HRC), such that the periodic fault impulse can be isolated and preserved in advance. Then, both LRC and HRC are, respectively, reconstructed using an online dictionary learning algorithm, i.e., Hankel K-singular value decomposition; thus, the reconstructed data can be obtained accordingly. Both simulation case and engineering case are applied to show the effectiveness and practicality of the proposed approach.

The large load capacity, compact size and high-power density have made planetary gearboxes widely applied in heavy machinery such as wind turbines. Although well designed, planetary gearboxes are vulnerable to fatigue crack because of harsh working environment such as high loads. Fatigue crack may eventually cause failures of planetary gearboxes if not detected early. This study investigates the dynamics of a planetary gear system with a crack in a sun gear tooth. This paper considers the gravity effect, carrier-ring bearings clearance, gyroscopic and centrifugal forces that reported studies have not considered together yet. Simulation results reveal that gravity produces a sinusoidal shift in the dynamic response of a planetary gearbox in the time domain. The carrier-ring bearing clearance causes crack-induced impulses to be more apparent. These results can help researchers develop more effective fault diagnosis strategies especially for early crack in the sun gear.

Planetary gearboxes are widely used in machinery such as wind turbines and helicopters. To maximize their effectiveness over their lifecycle, condition monitoring is often used, and proper health indexes can be developed utilizing condition monitoring data. A reported study treats health index development for electric motors as a regression problem and uses feedforward neural network to find the relationship between condition monitoring data and the health index. However, the reported method assumed a fixed step size following sequential ordering to select the optimal features. To improve the reported method, this paper proposes a feature selection strategy using genetic algorithm. A new health index (HI) is developed for planetary gearboxes. Experiments have shown that the proposed HI outperforms the reported HI. The proposed HI can better reflect the comprehensive health condition of planetary gearboxes.

Frequency band selection (FBS) in rotating machinery fault diagnosis aims to recognize frequency band location including a fault transient out of a full band spectrum, and thus fault diagnosis can suppress noise influence from other frequency components. Impulsiveness and cyclostationarity have been recently recognized as two distinctive signatures of a transient. Thus, many studies have focused on developing quantification metrics of the two signatures and using them as indicators to guide FBS. However, most previous studies almost ignore another aspect of FBS, i.e. health reference, which significantly affect FBS performance. To address this issue, this paper investigates importance of a health reference and recognize it as the third critical aspect in FBS. With help of the health reference, the frequency band where the fault transient exists could be located. A novel approach based on classification is proposed to integrate all three aspects (impulsiveness, cyclostationarity, and health reference) for FBS. Classification accuracy is developed as a novel indicator to select the most sensitive frequency band for rotating machinery fault diagnosis. The proposed method (coined by accugram) has been validated on benchmark and experiment datasets. Comparison results show its effectiveness and robustness over conventional envelope analysis, the kurtogram, and the infogram.

Background The contribution of HIV to COVID‐19 outcomes in hospitalized inpatients remains unclear. We conducted a multi‐centre, retrospective matched cohort study of SARS‐CoV‐2 PCR‐positive hospital inpatients analysed by HIV status. Methods HIV‐negative patients were matched to people living with HIV (PLWH) admitted from 1 February 2020 to 31 May 2020 up to a 3:1 ratio by the following: hospital site, SARS‐CoV‐2 test date ± 7 days, age ± 5 years, gender, and index of multiple deprivation decile ± 1. The primary objective was clinical improvement (two‐point improvement or better on a seven‐point ordinal scale) or hospital discharge by day 28, whichever was earlier. Results A total of 68 PLWH and 181 HIV‐negative comparators were included. In unadjusted analyses, PLWH had a reduced hazard of achieving clinical improvement or discharge [adjusted hazard ratio (aHR) = 0.57, 95% confidence interval (CI): 0.39–0.85, p = 0.005], but this association was ameliorated (aHR = 0.70, 95% CI: 0.43–1.17, p = 0.18) after additional adjustment for ethnicity, frailty, baseline hypoxaemia, duration of symptoms prior to baseline, body mass index (BMI) categories and comorbidities. Baseline frailty (aHR = 0.79, 95% CI: 0.65–0.95, p = 0.011), malignancy (aHR = 0.37, 95% CI 0.17, 0.82, p = 0.014) remained associated with poorer outcomes. The PLWH were more likely to be of black, Asian and minority ethnic background (75.0% vs 48.6%, p = 0.0002), higher median clinical frailty score [3 × interquartile range (IQR): 2–5 vs, 2 × IQR: 1–4, p = 0.0069), and to have a non‐significantly higher proportion of active malignancy (14.4% vs 9.9%, p = 0.29). Conclusions Adjusting for confounding comorbidities and demographics in a matched cohort ameliorated differences in outcomes of PLWH hospitalized with COVID‐19, highlighting the importance of an appropriate comparison group when assessing outcomes of PLWH hospitalized with COVID‐19.

• Background:Antiretroviral therapy (ART) has led to dramatic improvements in survival for people living with HIV, but is unable to cure infection, or induce viral control off therapy. Designing intervention trials with novel agents with the potential to confer a period of HIV remission without ART, remains a key scientific and community goal. We detail the rationale, design, and outcomes of a randomised, placebo-controlled trial of two HIV-specific long-acting broadly neutralising antibodies (bNAbs); 3BNC117-LS and 10-1074-LS, which target CD4 binding site and V3 loop respectively, on post-treatment viral control.• Methods:RIO is a randomised, placebo-controlled, double-blinded prospective phase II study. Eligible individuals will have started ART within 3 months of primary HIV infection and have viral sequences that appear to be sensitive to both bNAbs. It will randomise 72 eligible participants 1:1 to the following arms via a two-stage design. In stage 1, arm A participants are given dual long-acting (LS-variants) bNAbs infusions, followed by intensively monitored Analytical Treatment Interruption (ATI) (n=36); in arm B, participants receive placebo infusions followed by ATI. The primary endpoint will be time to viral rebound within 36 weeks after ATI. Upon viral rebound, the participant and researcher are unblinded. Participants in arm A recommence ART and complete the study. Participants in arm B are invited to restart ART and enroll into stage 2 where they will receive open-label LS bNAbs, followed by a second ATI 24 weeks after. Secondary and exploratory endpoints include adverse events, time to undetectable viraemia after re-starting ART, immunological markers, HIV proviral DNA, serum bNAb concentrations in blood, bNAb resistance at viral rebound, and quality of life measures.• Discussion:The two-stage design was determined in collaboration with community involvement. This design allows all participants the option to receive bNAbs. It also tests the hypothesis that bNAbs may drive sustained HIV-control beyond the duration of detectable bNAb concentrations. Community representatives were involved at all stages. This included the two-stage design, discussion on the criteria to restart ART, frequency of monitoring visits off ART and reducing the risk of onward transmission to HIV-negative partners. It also included responding to the challenges of COVID-19.Trial registration: The protocol is registered on Clinical.trials.gov and EudraCT and has approval from UK Ethics and MHRA.

Following changes in national antiretroviral therapy (ART) guidelines removing the CD4 threshold for initiation of ART, we evaluated the time to ART initiation and reasons for delayed or non-initiation. A retrospective notes review of 292 newly diagnosed HIV-positive individuals attending two London clinics between August 2015 and December 2016 was performed. Two hundred and fifty-four of 292 (87%) individuals started ART. Median time to ART initiation was 29 days (range: 0–514). Thirty of 292 (13%) did not start ART. Rates of virological suppression at six months were similar regardless of time to ART initiation. People who inject drugs (16.7% vs. 3.6%) (p = 0.009), having a higher median baseline CD4 cell count (500 vs. 388 cells/mm3, p = 0.001), and having gastrointestinal/liver co-morbidities (23% vs. 9%, p = 0.001) were associated with delayed ART initiation. The cohort not on ART had a higher median baseline CD4 cell count (500 vs. 388 cells/mm3, p

Background Atrial fibrillation is a paroxysmal heart disease without any obvious symptoms for most people during the onset. The electrocardiogram (ECG) at the time other than the onset of this disease is not significantly different from that of normal people, which makes it difficult to detect and diagnose. However, if atrial fibrillation is not detected and treated early, it tends to worsen the condition and increase the possibility of stroke. In this paper, P-wave morphology parameters and heart rate variability feature parameters were simultaneously extracted from the ECG. A total of 31 parameters were used as input variables to perform the modeling of artificial intelligence ensemble learning model. Results This paper applied three artificial intelligence ensemble learning methods, namely Bagging ensemble learning method, AdaBoost ensemble learning method, and Stacking ensemble learning method. The prediction results of these three artificial intelligence ensemble learning methods were compared. As a result of the comparison, the Stacking ensemble learning method combined with various models finally obtained the best prediction effect with the accuracy of 92%, sensitivity of 88%, specificity of 96%, positive predictive value of 95.7%, negative predictive value of 88.9%, F1 score of 0.9231 and area under receiver operating characteristic curve value of 0.911. Conclusion In feature extraction, this paper combined P-wave morphology parameters and heart rate variability parameters as input parameters for model training, and validated the value of the proposed parameters combination for the improvement of the model’s predicting effect. In the calculation of the P-wave morphology parameters, the hybrid Taguchi-genetic algorithm was used to obtain more accurate Gaussian function fitting parameters. The prediction model was trained using the Stacking ensemble learning method, so that the model accuracy had better results, which can further improve the early prediction of atrial fibrillation.

Background Neuropathic pain negatively affects quality of life among people living with HIV (PLWH). This study examined the feasibility of conducting a full-scale randomized-controlled trial of online acceptance and commitment therapy ("ACT OPEN") for neuropathic pain in PLWH. Methods Using a parallel-groups design, thirty-eight participants were randomized to ACT OPEN or a waitlist control (2:1). Participants completed standard self-report outcome measures at baseline, and two- and five-months post-randomization. Participants were aware of their allocation, but assessment was blinded. Results Twenty-five participants were randomized to ACT OPEN and 13 to the control (of 133 referrals). ACT OPEN completion was 69% and two-month trial retention was 82%. Treatment credibility and satisfaction scores for ACT OPEN were comparable to scores reported in previous trials of cognitive-behavioural treatments for pain. Four adverse events were reported during the study, including one serious adverse event; all of these were unrelated to the research procedures. Small to moderate effects and 95% confidence intervals suggest that the true effect may favour ACT OPEN for improvements in pain intensity/interference and depression. Conclusions A full-scale RCT of online ACT for pain management in PLWH may be feasible with refinements to trial design to facilitate recruitment. Significance Research on pain management in people living with HIV has primarily focused on pharmacological treatments with limited success. This is the first study to show the potential feasibility of a psychological treatment based on acceptance and commitment therapy delivered online and tailored for pain management in people with HIV ("ACT OPEN"). ACT OPEN may be a promising treatment in this population and further evaluation in a full-scale randomized-controlled trial appears warranted. Trial registration The trial was registered (clinicaltrials.gov; NCT03584412).

Deep learning methods have shown great potential to provide reliable remaining useful life (RUL) predictions in Prognostics and Health Management applications. However, deep learning models, particularly supervised learning methods, are strongly dependent on a large number of failure histories. In practice, engineering assets are generally replaced by new ones before failure during planned maintenance, resulting in a small number of failure histories and often times more than twice as many suspension histories. In this paper, a semi-supervised generative adversarial network (GAN) regression model is developed to consider both failure and suspension histories for RUL predictions. The proposed GAN model utilizes conditional multi-task objective functions to capture useful information from suspension histories to improve prediction accuracy, instead of simply treating them as unlabeled data. The method will not directly predict the failure times of suspension histories, but match the statistical information between similar failure and suspension histories to the greatest extent for model training. As a result, the failure information of suspension histories will not only rely on the failure histories but also on the generated data, thereby improving the model generalization, especially when the amount of data is limited. In addition, a robustness evaluation method is proposed to assess the uncertainty of the prognostic model caused by the scarce failure data. The accuracy and credibility of the proposed approach are validated by using two case studies.

This paper proposes a new feature extraction method using time–frequency image data for fault diagnosis of variable-speed rotating machinery. Time-frequency representation (TFR) is widely used to analyze time-varying behaviors of rotating machinery. Recently, methods have been developed to extract fault-related features from TFR image data. However, these methods can be only applied to in-phase TFR image data, or have limited sensitivity because they cannot utilize the characteristics of faults in rotating machinery. Therefore, the research outlined in this paper proposes a new fault feature for rotating machinery under variable-speed conditions. The proposed feature enhances sensitivity by exploiting faulty behaviors in the TFR image data. Two experimental case studies are presented to demonstrate the performance of the proposed method: a planetary gearbox and a spur gearbox. From the results, we conclude that the proposed method shows higher fault sensitivity than the previous image-based features, while showing consistent behavior under different phases of TFR image data.

Traditional intelligent fault diagnosis assumes that the training and testing samples are drawn from the same distribution. This assumption does not hold when working condition changes, as variable working condition can make the training and the testing datasets have different distributions. A novel working condition might be encountered in the testing stage, and there will be no label available under that novel working condition. This paper studies domain adaptation for gear crack level diagnosis under variable loading. A new two-stage fault diagnostic method for variable load condition is developed based on adversarial training strategy and gradient reversal layer. Both labeled and unlabeled data are utilized to learn best model for the novel load condition. An experimental case study is carried out to demonstrate the effectiveness of the proposed method.

Rotating machinery like wind turbines often operates under varying speed conditions. Measuring speed signals is critical for vibration-based condition monitoring of rotating machines. However, in real applications, sometimes it is difficult to install speed sensors to collect speed signals due to physical space and/or cost restrictions. Considering vibration signals are widely collected for condition monitoring, this paper proposes a virtual rotating speed meter which can extract the rotating speed of a rotating machine from its vibration signals. The proposed virtual speed meter is a framework of deep neural network in nature, which takes the vibration signal as the input, and directly outputs the corresponding speed signal. The training of this neural network is based on transfer learning across two machines: a source machine and a target machine. The source machine has massive historically measured vibration and speed signals available, while the target machine only has vibration available. The virtual speed meter is firstly trained with historical vibration signals and speed signals of the source machine, and then fine-tuned to fit a target machine to extract its speed from its vibration signals. Limited speed labels of the target machine for transfer learning are estimated using a reported signal processing method named Path Optimization. Case studies on two experimental datasets validate the effectiveness of the proposed virtual rotating speed meter.

Wind turbine power output estimation is an important problem in wind energy research work. Deterministic and probabilistic power curve models were reported when predicting the wind turbine power output. This paper proposes a probabilistic power curve model, and demonstrates it using field data from a wind farm in Alberta, Canada. Normal distribution and Weibull distribution are used to represent the probability density function of power output at various wind speed. Monte Carlo simulation is used to generate random predicting power output. The predicted result is compared with the observed data by 3 measurements and the proposed model is found performs better than other deterministic models and probabilistic models.

Gear wear is inevitable during the service life of gearboxes and may lead to catastrophic failure. As an important micro-level wear feature, tooth surface roughness directly affects the gear wear progression (lubrication regimes, wear mechanisms and wear rates) and lifespan of a gearbox. Therefore, it is important to monitor surface roughness changes. The tooth surface roughness induces random vibration signals with cyclic amplitude modulation. Reported works used an indicator of second-order cyclostationarity (ICS2) to assess such signals. However, the ICS2 gives a poor correlation with surface roughness. This paper presents the use of an Autoregressive Conditional Heteroskedasticity (ARCH) model to represent the random vibration signals with cyclic amplitude modulation. ARCH model parameter serves as an indicator to assess the changes in gear tooth surface roughness. A laboratory dataset was used to validate the effectiveness of the ARCH model in assessing surface roughness level. Results have shown that using the ARCH model returns a more accurate assessment result than the ICS2.

Planetary gearboxes are widely used in industrial applications, such as wind turbines, automobiles, and mining equipment. Although well designed, planetary gearboxes may experience failures if no early detection is conducted due to harsh working environment, such as high torque and heavy dust. To improve system reliability and reduce maintenance cost, it is important to carry out early fault detection. This study investigates the dynamic responses of a planetary gearbox with a tooth root crack in the sun gear. Reported studies have not considered the effects of bearing clearance on dynamic responses of planetary gearboxes with sun gear tooth crack. In this study, we consider the carrier bearing clearance, planet gear bearing clearance, and sun gear bearing clearance in a planetary gearbox with fixed ring gear. Simulation results reveal that if the bearing clearance is much smaller than the displacement of its corresponding gear, the clearance will not affect the dynamic response. Otherwise, the clearance will increase the displacement of its corresponding gear and change the dynamic responses of other gears.

Background: As exosomes have been confirmed as a reservoir of siRNAs involved in certain diseases, the current study aims to investigate whether exosomal-siRNA could exert a protective role in spinal cord injury (SCI). Methods and Results: Exosomes in our experiment were isolated from lysosomal membrane-associated protein 2b (Lamp2b) overexpression HEK 293T cells, and purity of exosomes was characterized by the expression of CD9, CD47, and CD63 via western blot. Furthermore, the siRNA pool contains four siRNAs including siRNA-NgR, siRNA-LINGO-1, siRNA-Troy, and siRNA-PTEN was loaded to the exosomes, which indicated a significant role for the siRNA pool in reducing the expression of axon growth inhibitory factors. Upon the completion of loading into exosomes (exo-siRNA pool), the exo-siRNA pool was injected into primary cortical neurons of the SCI model in rats before cell proliferation and Rho expression were determined With the results revealed that purified addition could be applied to future experiments. The exo-siRNA pooled transfection caused downregulation of axon growth suppressors in primary cortical neurons including Nogo receptors (NgR), leucine-rich repeats and immunoglobulin domain-containing protein 1 (LINGO-1), Troy, and phosphatase and tenson homolog (PTEN). Cell proliferation and Rho expression of primary cortical neurons inhibited the expression of axonal growth inhibitors in rats with SCI by transfecting exogenous Sirna. Conclusion: This study confirmed that exosomes derived from Lamp2b overexpression HEK 293T cells facilitated both the recovery of functions and the survival of neurons when being loaded with the siRNA pool.

Late HIV diagnosis is associated with significant mortality in people living with HIV (PLWH) and high numbers of missed opportunities (MO) for earlier testing have been identified. A pilot of a national late diagnosis review process (LDRP) was undertaken in 15 HIV services evaluating the feasibility of LDRP implementation, as a patient safety initiative. All newly diagnosed PLWH with CD4 counts

The use of digital vending machines (VMs) to delivery HIV self-testing (HIVST) could expand HIV testing in priority populations. We surveyed primarily Black African (BA) participants and other minority ethnicities, to identify acceptability, preferences, and concerns of using VMs for HIVST dispensing. A structured survey was developed with Black African and Caribbean, Latin American and other Minorities (BLAM) communities, and distributed between September 2018 and January 2019. Participants were recruited using mobile tablet surveys distributed by outreach volunteers, and online through BLAM communities’ websites, workshops, and language-specific messages on social media. Descriptive analyses were undertaken stratified by ethnic groups. One hundred and twenty-eight (67.0%) participants identified as BAs, 31 (16.2%) Black Caribbeans (BCs), 22 (11.5%) Latin Americans (LAs), and 10 (5.2%) other non-white ethnicities (ONWEs). Rates of willingness to use the HIVST were high in all groups except BCs (BAs 77.9%, BCs 53.6%, LAs 81.8%, ONWEs 80.0%). Most participants favoured healthcare-associated venues for VM placement, but there were differences in community venues favoured by different ethnic groups and concerns reported. HIVST is acceptable in many BLAM communities and increases understanding of the concerns and how to address them in the design of VMs for HIVST, to expand HIV testing in these priority communities.

To achieve planetary gearbox fault classification, vibration signal analysis has been widely employed with rich information about the health status and easy measurement. It is critical to extract features with enough health status information for fault classification. The self-adaptation of ensemble empirical mode decomposition (EEMD) indicates the dependence between the raw vibration signal and EEMD-decomposed intrinsic mode functions (IMFs). In this study, we develop a novel fault feature vector based on the dependence. To develop the dependence-based feature vector, simulated vibration signals with different sun gear tooth crack levels are analyzed. The dependence between the raw signal and each IMF is investigated by Archimedean copulas. With the goodness-of-fit test, the copula estimation closest to the perfect fit is selected for dependence representation. The parameter of the selected copula is applied to develop the dependence-based feature vector. To test the ability of the dependence-based feature vector in fault classification for a real planetary gearbox, experimental vibration signals with different gear fault levels at different gears are classified by a multi-class support vector machine. The classification accuracy of the developed feature vector is compared with that of a reported indicator. Results show the dependence-based feature vector provides higher classification accuracy than the reported, indicating the developed feature vector contains more health status information. The developed feature vector can serve better for planetary gearbox fault classification.

Time-varying mesh stiffness is one of the main internal excitation sources of gear dynamics. Accurate evaluation of gear mesh stiffness is crucial for gear dynamic analysis. This study is devoted to developing new models for spur gear mesh stiffness evaluation. Three models are proposed. The proposed model 1 can give very accurate mesh stiffness result but the gear bore surface must be assumed to be rigid. Enlighted by the proposed model 1, our research discovers that the angular deflection pattern of the gear bore surface of a pair of meshing gears under a constant torque basically follows a cosine curve. Based on this finding, two other models are proposed. The proposed model 2 evaluates gear mesh stiffness by using angular deflections at different circumferential angles of an end surface circle of the gear bore. The proposed model 3 requires using only the angular deflection at an arbitrary circumferential angle of an end surface circle of the gear bore but this model can only be used for a gear with the same tooth profile among all teeth. The proposed models are accurate in gear mesh stiffness evaluation and easy to use. Finite element analysis is used to validate the accuracy of the proposed models.

Muscle damage after 30 maximal eccentric contractions of the elbow flexors (30MVEC) is reduced when the same exercise is performed by the opposite arm, and when two maximal voluntary isometric contractions at a long muscle length (2MVIC) are performed prior to 30MVEC by the same arm. This study investigated the hypothesis that 2MVIC would attenuate muscle damage after 30MVEC performed by the opposite arm. Untrained young (20-25 years) men were placed into 1 of 4 experimental groups that performed 2MVIC at 1 (1d), 2 (2d), 4 (4d), or 7 days (7d) before 30MVEC by the opposite arm, or one control group that performed 30MVEC only (n = 13/group). Changes in indirect muscle damage markers after 30MVEC were compared among the groups by mixed-design two-way ANOVA. Maximal voluntary concentric contraction torque, range of motion, plasma creatine kinase activity, and muscle soreness did not change significantly after 2MVIC. Changes in these variables after 30MVEC were smaller (P

Gearbox is widely used in industrial and military applications. Due to high service load, harsh operating conditions or inevitable fatigue, faults may develop in gears. If the gear faults cannot be detected early, the health will continue to degrade, perhaps causing heavy economic loss or even catastrophe. Early fault detection and diagnosis allows properly scheduled shutdowns to prevent catastrophic failure and consequently result in a safer operation and higher cost reduction. Recently, many studies have been done to develop gearbox dynamic models with faults aiming to understand gear fault generation mechanism and then develop effective fault detection and diagnosis methods. This paper focuses on dynamics based gearbox fault modeling, detection and diagnosis. State-of-art and challenges are reviewed and discussed. This detailed literature review limits research results to the following fundamental yet key aspects: gear mesh stiffness evaluation, gearbox damage modeling and fault diagnosis techniques, gearbox transmission path modeling and method validation. In the end, a summary and some research prospects are presented.

Failure mode and effects analysis (FMEA) is a structured method used during a given stage of the system life cycle to understand all probable failure modes and the effects of their occurrences. The risk priority number (RPN) is calculated in FMEA to select more critical failure modes by multiplying three factors: occurrence, detection, and severity. In the literature, these three factors are defined qualitatively without any underlying model, and multiple definitions and conflicting interpretations exist for each factor. As the interrelationships between the three RPN factors are not known, previous research has treated each factor as a criterion in multiple criteria decision making, under the assumption that the three factors are independent of each other. In this paper, we present a general model to explain the functional relationship among the three factors. Using the model, we discuss the unique role of each factor for comparing the risk of different failure modes.

Planet bearing vibration signals are highly complex due to intricate kinematics (involving both revolution and spinning) and strong multiple modulations (including not only the fault induced amplitude modulation and frequency modulation, but also additional amplitude modulations due to load zone passing, time-varying vibration transfer path, and time-varying angle between the gear pair mesh lines of action and fault impact force vector), leading to difficulty in fault feature extraction. Rolling element bearing fault diagnosis essentially relies on detection of fault induced repetitive impulses carried by resonance vibration, but they are usually contaminated by noise and therefor are hard to be detected. This further adds complexity to planet bearing diagnostics. Spectral negentropy is able to reveal the frequency distribution of repetitive transients, thus providing an approach to identify the optimal frequency band of a filter for separating repetitive impulses. In this paper, we find the informative frequency band (including the center frequency and bandwidth) of bearing fault induced repetitive impulses using the spectral negentropy based infogram. In Fourier spectrum, we identify planet bearing faults according to sideband characteristics around the center frequency. For demodulation analysis, we filter out the sensitive component based on the informative frequency band revealed by the infogram. In amplitude demodulated spectrum (squared envelope spectrum) of the sensitive component, we diagnose planet bearing faults by matching the present peaks with the theoretical fault characteristic frequencies. We further decompose the sensitive component into mono-component intrinsic mode functions (IMFs) to estimate their instantaneous frequencies, and select a sensitive IMF with an instantaneous frequency fluctuating around the center frequency for frequency demodulation analysis. In the frequency demodulated spectrum (Fourier spectrum of instantaneous frequency) of selected IMF, we discern planet bearing fault reasons according to the present peaks. The proposed spectral negentropy infogram based spectrum and demodulation analysis method is illustrated via a numerical simulated signal analysis. Considering the unique load bearing feature of planet bearings, experimental validations under both no-load and loading conditions are done to verify the derived fault symptoms and the proposed method. The localized faults on outer race, rolling element and inner race are successfully diagnosed.

Intelligent fault diagnosis of machines has received much attention in this big data era. Most reported models are constructed under the assumption that the training and testing data are from the same distribution. However, data distribution will shift due to working condition changes, posing challenges on the performance of intelligent models. This study considers the case that out of many known working conditions with labeled historical data, the model is to be used under another unlabelled target working condition. A multiple source domain adaptation method is proposed to learn fault-discriminative but working-condition-invariant features from raw vibration signals. Different known working conditions will be assigned with different weights, on the basis of their distributional similarities to the target working condition. Two case studies are carried out to validate the effectiveness of the proposed method, respectively on rotating speed changes and load level changes.

Driven by the potential applications of sliding bearings in planetary gearboxes for wind turbines, the wear prognosis of heavy loaded sliding bearings under low rotational speeds is an important aspect. The aims of this study are to identify an adequate condition monitoring technique and demonstrate the potential of data-driven wear monitoring for scenarios, where transient wear data for data-driven monitoring is not available. In a first step, Acoustic Emission (AE) technique has been applied to a special test rig for planetary gearbox sliding bearings. It was demonstrated that AE can be used to distinguish between wear-critical mixed friction and hydrodynamic regime. In a second step, a data-driven method for wear monitoring was developed and applied to a component test rig for sliding bearings. For validation and generation of condition monitoring data, sliding bearings from the same bronze material were subjected to steady speed conditions in mixed lubrication regime with different loads as well as runtime. The results from these experiments and results from validated physical wear simulations are the input parameters for the proposed data-driven approach. The wear prognosis is performed with recurrent neural networks, which can predict the transient degradation. With the developed data-driven approach to wear monitoring, a good accuracy can be achieved that is capable of real-time wear monitoring.

In industry (e.g., wind power), gearboxes often operate under random speed variations. A condition monitoring system is expected to detect faults and assess their severity using vibration signals collected under different speed profiles. A few studies have been reported for condition monitoring of gearboxes under random speed variations, including a novelty diagnostic method and a support vector machine (SVM) based method. However, these methods either are based on the strict assumption that the rotating speed does not vary significantly within a rotating cycle or have the drawback of low classification accuracy. This paper presents a time series model-based method for gear tooth crack detection and severity assessment under random speed variation. Specifically, the rotating speed and phase are considered as covariates in a linear parameter varying autoregression (AR) model for representing impulsive vibration signals. We propose refined B-splines for mapping the dependency between AR coefficients and the rotating phase. The performance of the presented time series model-based method has been validated using laboratory signals. The presented method can assess 93.8% of the tooth crack severity state correctly, which is better than the novelty diagnostic method (74.4%) and SVM-based method (87.7%).

Planetary gearbox is a critical component for rotating machinery. It is widely used in wind turbines, aerospace and transmission systems in heavy industry. Thus, it is important to monitor planetary gearboxes, especially for fault diagnostics, during its operational conditions. However, in practice, operational conditions of planetary gearbox are often characterized by variations of rotational speeds and loads, which may bring difficulties for fault diagnosis through the measured vibrations. In this paper, phase angle data extracted from measured planetary gearbox vibrations is used for fault detection under non-stationary operational conditions. Together with sample entropy, fault diagnosis on planetary gearbox is implemented. The proposed scheme is explained and demonstrated in both simulation and experimental studies. The scheme proves to be effective and features advantages on fault diagnosis of planetary gearboxes under non-stationary operational conditions.

Tooth pitting is a common failure mode of a gearbox. Many researchers investigated dynamic properties of a gearbox with localized pitting damage on a single gear tooth. The dynamic properties of a gearbox with pitting distributed over multiple teeth have rarely been investigated. In this paper, gear tooth pitting propagation to neighboring teeth is modeled and investigated for a pair of spur gears. Tooth pitting propagation effect on time-varying mesh stiffness, gearbox dynamics and vibration characteristics is studied and then fault symptoms are revealed. In addition, the influence of gear mesh damping and environmental noise on gearbox vibration properties is investigated. In the end, 114 statistical features are tested to estimate tooth pitting growth. Statistical features that are insensitive to gear mesh damping and environmental noise are recommended.

Time Limited Dispatch (TLD) allows a degraded redundant system to operate for a limited length of time before it is sent for repair. Existing TLD analysis methods do not consider dispatch reliability requirement as a constraint, thus unacceptable flight delays or cancellations may be experienced when aircraft are dispatched following the intervals determined by existing methods. In this paper, an improved TLD analysis method with constraints of both dispatch reliability and average safety levels is proposed based on Markov process analysis. Dispatch strategies for multiple fault TLD models are proposed, the average safety level is redefined, and both the dispatch reliability and the average safety level are expressed by dispatch intervals based on the Markov processes. Finally, examples illustrate the advantages of the proposed method. The results show that the requirements for both dispatch reliability and average safety level are ensured when aircraft are dispatched following the intervals determined by the proposed method.

Safety instrumented system (SIS) is widely applied to reduce or prevent risk in industry. Practical SIS commonly meets with epistemic uncertainty arising from incompleteness of knowledge on various input parameters due to lack of data. Epistemic uncertainty can be reduced by collecting more knowledge or data. Epistemic uncertainty in input parameters can lead to variation in probability of failure to perform its intended functions on demand (PFD) for an SIS. This paper employs the complementary cumulative distribution function of PFD to define exceedance probability (EP) that the PFD exceeds a prescribed value. Sensitivity analysis is further investigated, analyzing the effect of an epistemically uncertain input parameter with truncated distribution on EP of an SIS. We have derived the analytic expression for evaluating the effect, and only an evaluation is needed to estimate the effects of all the parameters. Two examples are employed to demonstrate the applicability of the proposed method. We further compare the effects for the truncated and nontruncated parameter situations. Their results converge to the same ones as the truncated region of an input parameter decreases to zero. When the truncated region of an input parameter is less than 10–3 , the corresponding results can be approximated by the ones of the nontruncated case with lower computational cost.

This paper presents a novel signal processing scheme, namely time-varying morphological filtering (TMF), for rolling element bearing fault detection. In contrast with the multiscale morphological filtering (MMF) method, the structure element (SE) used in TMF is no longer fixed. It adjusts adaptively according to the extreme points of a signal so that the raw signal can be fit more accurately. In addition, the MMF needs to execute morphological operations multiple times, whereas the TMF can finish the filtering in one time operation. Consequently, TMF has a significant advantage in terms of computation efficiency. Experimental vibration signals collected from a bearing test rig are employed to evaluate the effectiveness of TMF. The results show that the proposed method can extract fault features of defective rolling element bearings with high computational efficiency. Moreover, five SEs are compared for TMF. The results show that the Dolph-Chebyshev SE performs best.

Complex signal analysis is a key topic in machinery fault diagnosis. For complex multi-component signals of various morphological contents, the commonly used basis expansion based signal processing method lacks adaptability and flexibility, thus being ineffective to extract the embedded meaningful information. Sparse signal representation has excellent adaptability and high flexibility in describing arbitrary complex signals based on atomic decomposition over redundant and over-complete dictionary, thus being free from the limitations imposed by orthogonal basis, and providing an effective approach to feature extraction from intricate signals for machinery fault diagnosis. This paper presents a systematic review on sparse signal representation, especially on two key topics, i.e. atomic decomposition algorithms (such as greedy pursuit, lp norm regularization and iterative shrinkage/thresholding) and dictionary design methods (including analytic dictionary design and dictionary learning) reported in more than 70 representative articles published since 1990. Their fundamental principles, advantages and disadvantages, and applications to machinery fault diagnosis, are examined. Some examples are provided to illustrate their performance.

Local mean decomposition (LMD) has been developed for modulation information mining. Its performance is dependent on boundary condition, envelope estimation, and stopping criterion. This paper proposes a soft sifting stopping criterion that enables LMD to achieve a self-adaptive stop for each sifting process. In the proposed method, we define an objective function that considers two characteristics, namely, the root mean square and the excess kurtosis, of the target signal. To optimize this objective function, a heuristic mechanism is proposed to automatically determine the optimal number of sifting iterations. Experimental results on simulated signals demonstrate the effectiveness of the proposed soft sifting stopping criterion for improving the accuracy of LMD, and finally the proposed method is applied to modulation information mining for gear fault diagnosis.

Effective signal processing methods are essential for machinery fault diagnosis. Most conventional signal processing methods lack adaptability, thus being unable to well extract the embedded meaningful information. Adaptive mode decomposition methods have excellent adaptability and high flexibility in describing arbitrary complicated signals, and are free from the limitations imposed by conventional basis expansion, thus being able to adapt to the signal characteristics, extract rich characteristic information, and therefore reveal the underlying physical nature. This paper presents a systematic and up-to-date review on adaptive mode decomposition in two major topics, i.e., mono-component decomposition algorithms (such as empirical mode composition, local mean decomposition, intrinsic time-scale decomposition, local characteristic scale decomposition, Hilbert vibration decomposition, empirical wavelet transform, variational mode decomposition, nonlinear mode decomposition, and adaptive local iterative filtering) and instantaneous frequency estimation approaches (including Hilbert-transform-based analytic signal, direct quadrature, and normalized Hilbert transform based on empirical AM-FM decomposition, as well as generalized zero-crossing and energy separation) reported in more than 80 representative articles published since 1998. Their fundamental principles, advantages and disadvantages, and applications to signal analysis in machinery fault diagnosis, are examined. Examples are provided to illustrate their performance.

When there is a single cracked tooth in a planet gear, the cracked tooth is enmeshed for very short time duration in comparison to the total time of a full revolution of the planet gear. The fault symptom generated by the single cracked tooth may be very weak. This study aims to develop a windowing and mapping strategy to interpret the vibration signal of a planetary gear at the tooth level. The fault symptoms generated by a single cracked tooth of the planet gear of interest can be extracted. The health condition of the planet gear can be assessed by comparing the differences among the signals of all teeth of the planet gear. The proposed windowing and mapping strategy is tested with both simulated vibration signals and experimental vibration signals. The tooth signals can be successfully decomposed and a single tooth fault on a planet gear can be effectively detected.

Time-varying mesh stiffness is one of the main internal excitations of gear dynamics. With the growth of gear tooth fault, like pitting, cracking and spalling, the mesh stiffness amplitude changes and consequently the dynamic properties of the gear system change. This study is devoted to deriving equations of the mesh stiffness of a pair of external spur gears with tooth pitting. Different pitting severity levels are modeled. The influence of tooth pitting on the gear mesh stiffness is investigated. The relationship between pitting severity and mesh stiffness is established. The proposed method is validated to be effective in mesh stiffness evaluation by comparing with a finite element model.

Batten disease, or juvenile NCL, is a fatal neurodegenerative disorder that occurs due to mutations in the CLN3 gene. Because the function of CLN3 remains unclear, experimental therapies for JNCL have largely concentrated upon the targeting of downstream pathomechanisms. Neuron loss is preceded by localized glial activation, and in this proof-of-concept study we have investigated whether targeting this innate immune response with ibuprofen in combination with the neuroprotective agent lamotrigine improves the previously documented beneficial effects of immunosuppressants alone. Drugs were administered daily to symptomatic Cln3 -/- mice over a 3 month period, starting at 6 months of age, and their impact was assessed using both behavioral and neuropathological outcome measures. During the treatment period, the combination of ibuprofen and lamotrigine significantly improved the performance of Cln3 -/- mice on the vertical pole test, slowing the disease-associated decline, but had less of an impact upon their rotarod performance. There were also moderate and regionally dependent effects upon astrocyte activation that were most pronounced for ibuprofen alone, but there was no overt effect upon microglial activation. Administering such treatments for longer periods will enable testing for any impact upon the neuron loss that occurs later in disease progression. Given the partial efficacy of these treatments, it will be important to test further drugs of this type in order to find more effective combinations.

Liver toxicity is a frequent adverse event following antiretroviral therapy for HIV, ranging from asymptomatic increases in liver enzymes to fulminant liver failure.1 2 Rilpivirine is a second-generation non-nucleoside reverse transcriptase inhibitor with low rates of liver enzyme elevations.3 We report a case of acute hepatitis secondary to rilpivirine with histological documentation. A 27-year-old Mediterranean man was diagnosed with HIV, with a baseline CD4 count of 0.359 ×109/L and an HIV-1 viral load of 21 525 copies/mL. His last negative HIV test was 2 years prior. His medical history included depression, G6PD deficiency and a beta thalassaemia carrier status. He reported minimal alcohol intake. His baseline liver markers were normal. Abacavir/lamivudine 600/300 mg once daily and raltegravir 400 mg two times per day were started on week 4 following diagnosis. Raltegravir was switched to rilpivirine 25 mg once …

Localized tooth crack in gearboxes may be reflected in impulse components of gearbox vibration signals. Crack induced impulses have been used for crack detection and fault diagnosis. In reported studies, researchers have used statistical indicators of the identified impulses, such as root mean square (RMS) and kurtosis, to track the growth of crack. These reported statistical indicators are only effective when crack levels are high and they are unable to detect tooth crack and assess crack severity in the early stage of crack propagation. In addition, no reported studies have focused on studying how tooth crack level affects crack induced impulses. Specifically, what the dominant segments of crack induced impulses are and which segment is affected more by crack growth within a certain crack level range. This paper uses dynamic modeling to study how crack level affects crack induced impulses. First, impulses are generated with a spur gearbox dynamic model under constant working conditions. Second, an exponentially damped sinusoidal model is utilized to fit the impulses and the Matrix Pencil Method is used for model parameter estimation. Finally, relationships between crack level and impulses are studied based on the obtained model parameters. The results have shown that the segments in the fifth and the sixth frequency bands of impulses are two dominant segments, while other segments have little contribution, for the gearbox system under investigation. Within a certain crack level range, there exists an impulse segment which is more affected by the crack level. In terms of the early stage of crack propagation, the segment in the sixth frequency band of the impulse is more affected by crack growth. On this basis, three new statistical indicators have been developed with the segment in sixth frequency band of the impulse and have shown their effectiveness for tooth crack severity level assessment in the early stage of crack propagation. These results have good potential for detection and severity assessment of early tooth cracks in gearboxes.

Due to the limited field of view of the pinhole camera, there is a lack of stability and accuracy in camera pose estimation applications such as visual SLAM. In order to solve this problem, a visual SLAM algorithm based on a panoramic camera is proposed to realize high-precision camera pose estimation and 3D sparse feature map construction. The indoor and outdoor experiments show that the proposed visual SLAM system is superior to that based on pinhole camera in both stability and precision, and it provides and effective solution for the localization and navigation of the mobile robots.

Rotating machines are widely used in industry and often work under harsh and varying speed conditions. Fault diagnosis under varying speed conditions is needed to prevent major shutdowns. This paper aims to develop an intelligent rotating machinery fault diagnosis strategy based on deep neural networks (DNNs) and order tracking (OT). The developed strategy can automatically conduct rotating machinery fault diagnosis under both constant and varying speed conditions. Case studies on a rolling element bearing dataset and a fixed-shaft gearbox dataset show the superiority in diagnosis accuracy of the proposed strategy over reported approaches.

In real-world applications, fault detection and diagnosis of planetary gearboxes are vital if it can be employed to avert catastrophic failure consequences in rotating machinery. Fault diagnosis usually starts with collecting vibration signals from rotating machinery. These vibration signals are usually produced in non-stationary operating conditions with time-varying loads and speeds, which makes fault diagnosis more challenging. Signal processing methods are typically selected for fault diagnosis to capture either time, frequency, or time-frequency based diagnostic features from measured vibration signals. Yet, it is usually a costly or time-consuming process and, sometimes, heavily dependent on human expertise. Although current deep learning algorithms offer an efficient and intelligent diagnostic strategy for fault diagnosis, unfortunately, most of the reported algorithms are basically only valid for the stationary operating conditions. To address the challenges of non-stationary operating conditions, in this paper, an Automatic Speed Adaption Neural Network (ASANN) model within the incorporation of instantaneous rotating speed is proposed, and it provides an end-to-end learning fashion with the guidance of rotating speed information. With the incorporating of instantaneous rotating speed information, the proposed ASANN model enables the extraordinary capacity for planetary gearbox fault detection under varying operational scenarios. The validity of the ASANN model is verified by an experimental investigation of fault diagnosis in a planetary gearbox.

Long-term reliable health condition monitoring (HCM) of a wind turbine is an essential method to avoid catastrophic failure results. Existing unsupervised learning methods, such as auto-encoder (AE) and de-noising auto-encoder (DAE) models, are utilized to the condition monitoring of wind turbines. The critical bottleneck of these models for monitoring is to determine a threshold for identifying different health conditions. Unfortunately, the threshold is usually set up with different kinds of calculation methods or even based on experience. Therefore, the uncertainty of the threshold will inevitably influence the accuracy of the monitoring process and may lead to misdiagnoses. To overcome this limitation, this research introduces a threshold self-setting HCM scheme, based on deep convolutional generative adversarial networks (DCGAN) and employed for defining a self-setting threshold to monitor wind turbine generator bearings. A threshold for HCM can be automatically created through the output of the G network in the DCGAN model, and the challenging problem of setting up a threshold can be solved. Besides, the use of Nash Equilibrium for training enables this scheme to become self-defined evaluators with a high level of consistency, without any human intervention and can be treated as a self-defined threshold, and it is a model self-tuning process. Furthermore, a sample discrepancy analysis based on the output of the G network is utilized so that a quantitative indicator of the fault severity in wind turbine generator bearings are provided. By tracking a real wind turbine dataset from the LU NAN wind farm in China, the effectiveness of the proposed scheme is verified.

The vibration signals of planet bearings feature more complex amplitude modulation and frequency modulation (AM-FM) than those of fixed-axis bearings, due to the effects of time-varying vibration transfer path, load zone passing and time-varying angles between gear pair mesh lines of action and impact force vector, in addition to that of bearing fault. Their Fourier spectra have a complex sideband structure, leading to difficulty in feature extraction and fault diagnosis. In order to address this issue, a joint amplitude and frequency demodulation analysis method is proposed to reveal the fault features, by considering the modulation characteristics. To thoroughly understand planet bearing vibration characteristics, the explicit equations for amplitude and frequency demodulated spectra of outer race, rolling element and inner race fault cases are derived, and the fault symptoms are summarized respectively. The signal is firstly decomposed into intrinsic mode functions (IMFs) by empirical mode decomposition (EMD), thus meeting the mono-component requirement by instantaneous frequency calculation. Then a sensitive component with the instantaneous frequency fluctuating around the resonance frequency is chosen for further frequency demodulation analysis. Finally, planet bearing fault can be diagnosed by matching the peaks identified in amplitude and frequency demodulated spectra with the theoretical fault characteristic frequencies. The proposed method is validated with both numerical simulated and lab experimental signal analyses.