Data-driven diagnosis of high temperature PEM fuel cells based on the electrochemical impedance spectroscopy: Robustness improvement and evaluation

Utilizing machine learning techniques for data-driven diagnosis of high temperature PEM fuel cells is beneficial and meaningful to the system durability. Nevertheless, ensuring the robustness of diagnosis remains a critical and challenging task in real application. To enhance the robustness of diagnosis and achieve a more thorough evaluation of diagnostic performance, a robust diagnostic procedure based on electrochemical impedance spectroscopy (EIS) and a new method for evaluation of the diagnosis robustness was proposed and investigated in this work. To improve the diagnosis robustness: (1) the degradation mechanism of different faults in the high temperature PEM fuel cell was first analyzed via the distribution of relaxation time of EIS to determine the equivalent circuit model (ECM) with better interpretability, simplicity and accuracy;(2) the feature extraction was implemented on the identified parameters of the ECM and extra attention was paid to distinguishing between the long-term normal degradation and other faults;(3) a Siamese Network was adopted to get features with higher robustness in a new embedding. The diagnosis was conducted using 6 classic classification algorithms—support vector machine (SVM), K-nearest neighbor (KNN), logistic regression (LR), decision tree (DT), random forest (RF), and Naive Bayes employing a dataset comprising a total of 1935 collected EIS. To evaluate the robustness of trained models: (1) different levels of errors were added to the features for performance evaluation;(2) a robustness coefficient (Roubust_C) was defined for a quantified and explicit evaluation of the diagnosis robustness. The diagnostic models employing the proposed feature extraction method can not only achieve the higher performance of around 100% but also higher robustness for diagnosis models. Despite the initial performance being similar, the KNN demonstrated a superior robustness after feature selection and re-embedding by triplet-loss method, which suggests the necessity of robustness evaluation for the machine learning models and the effectiveness of the defined robustness coefficient. This work hopes to give new insights to the robust diagnosis of high temperature PEM fuel cells and more comprehensive performance evaluation of the data-driven method for diagnostic application.

Formation of precipitates and recrystallization resistance in Al-Sc-Zr alloys

Al-x%Sc-0.11%Zr alloys （x=0, 0.02, 0.05, 0.08, 0.11, 0.15） were produced by a chill cast procedure. The effect of Sc content on the precipitation of Al3（Sc,Zr） during heat treatment at 475 °C for 12 h was studied. Nucleation, precipitation and distribution of Al3（Sc,Zr） precipitates were found to be strongly related to the Sc content. With increasing the Sc content, the average radius of the precipitates decreases, while the number density of the precipitates increases, as investigated by transmission electron microscopy （TEM）. The distribution of the precipitates becomes more and more homogeneous when the Sc content is increased. The recrystallization resistance of samples that was 90% cold rolled and isothermal annealed for half an hour in the temperature range of 200-600 °C was investigated. The results show that the recrystallization temperature varies from 250 °C for the alloy without Sc to about 600 °C for the alloy containing 0.15% Sc because of the high density of Al3（Sc,Zr） precipitates.

Precipitation behaviour of Al_3Zr precipitate in Al-Cu-Zr and Al-Cu-Zr-Ti-V alloys

The precipitation behaviours of Al3Zr precipitate in the Al-Cu-Zr and Al-Cu-Zr-Ti-V alloys were studied by transmission electron microscopy. Metastable Al3Zr precipitates are homogeneously nucleated in dendrite centres resulting in homogeneous distribution. However, the precipitation in the interdendritic regions is complex and the precipitation morphologies, helical-like and stripe-like shapes, were observed, which are composed of many spherical Al3Zr precipitates. The stripe-like precipitate clusters have preferential orientations along with the -100- Al directions, which is inferred to be related to θ′（Al2Cu） and θ phases. Addition of Cu can accelerate the L12→D023 structural transformation of the Al3Zr precipitate.

Effects of Sr^(2+) substitution on the structural, dielectric, and piezoelectric properties of PZT-PMN ceramics

This study described the structural, dielectric, and piezoelectric behavior of Pb1-xSrx[（Zr0.52Ti0.48）0.95（Mn1/3Nb2/3）0.05]O3 ceramics （PSZT-PMN, x=0, 0.025, 0.050, and 0.075）, prepared by a semi-wet route. X-ray diffraction, dielectric, and piezoelectric investigations were carried out to analyze the crystal structure. The relative dielectric constant and dielectric loss were both calculated as the functions of temperature. The room-temperature dielectric constant reaches a maximum for a Sr2＋-modified PZT-PMN ceramic with an x value of 0.050, which corresponds to the morphotropic phase boundary （MPB）. Raman spectroscopy studies also confirm the existence of this MPB for x=0.050. The piezoelectric strain coefficients （d33） value shows a maximum response for this composition. In addition, the phase transition temperature decreases significantly when the Sr2＋concentration increases in the PZT-PMN ceramics.

Technological Advances in Accelerated Wound Repair and Regeneration

We reviewed a number of wound repair, keloid and hypertrophic scar research methods that included lasers, microcurrent and ultra-low energy technologies. Laser research reports short-term improvement in wounds, keloid and hypertrophic scars, but without follow up to control for reoccurrence of keloids or diabetic lesions which generally reoccur following laser treatments. The microcurrent and ultra-low energy studies demonstrate significant healing where age is not a factor with no reoccurrence of diabetic wounds and other skin lesions. Our randomized, double-blind longitudinal research on eight wound repair clinical cases with an age range of 28 - 86, followed for one year, evidenced accelerated healing and no reoccurrence. The number of treatments required for substantial healing depended on the chronicity and severity of the lesion, with chronic severe lesions requiring more treatments, rather than age, a conclusion supported by ultra-low microcurrent research. These results on age-independent wound healing directly contradict a large body of literature postulating that healing is much slower with age due to immune insufficiency, age-accumulated oxidative stress, disrupted cell communications and sustained inflammation.

Dielectric, piezoelectric, and ferroelectric properties of lanthanum-modified PZTFN ceramics

Specimens of Pb1-1.5xLax（Zr0.53 Ti0.47）1-y-zFeyNb2O3 （x = 0, 0.004, 0.008, 0.012, and 0.016, y = z = 0.01） （PZTFN） ceramics were synthesized by a semi-wet route. In the present study, the effect of La doping was investigated on the structural, microstructural, dielectric, piezoelectric, and ferroelectric properties of the ceramics. The results show that, the tetragonal （space group P4mm） and rhombohedral （space group R3c） phases are observed to coexist in the sample at x = 0.012. Microstructural investigations of all the samples reveal that La doping inhibits grain growth. Doping of La into PZTFN improves the dielectric, ferroelectric, and piezoelectric properties of the ceramics. The hys- teresis loops of all specimens exhibit nonlinear behavior. The dielectric, piezoelectric and ferroelectric properties show a maximum response atx 〉 0.012, which corresponds to the morphotropic phase boundary （MPB）.

Impedance and ferroelectric properties of Sr^(2+) modified PZT-PMN ceramics

Sr^2＋ modified polycrystalline PZT-PMN ceramics were synthesized by a semi-wet route. Impedance spectroscopy studies indicate the bulk and grain boundary effects of PZT-PMN material along with the negative temperature coefficient of resistance. The bulk conductiv-ity exhibits an Arrhenius-type thermally activated hopping process which is supported by the AC conductivity behavior as a function of fre-quency and temperature. It is observed that the remnant polarization increases with an increase in the Sr2＋content in PZT-PMN.

Deep Learning in Medical Imaging and Drug Design

Over the last decade,deep learning(DL)methods have been extremely successful and widely used in almost every domain.Researchers are now focusing on the convergence of medical imaging and drug design using deep learning to revolutionize medical diagnostic and improvement in the monitoring from response to therapy.DL a new machine learning paradigm that focuses on learning with deep hierarchical models of data.Medical imaging has transformed healthcare science,it was thought of as a diagnostic tool for disease,but now it is also used in drug design.Advances in medical imaging technology have enabled scientists to detect events at the cellular level.The role of medical imaging in drug design includes identification of likely responders,detection,diagnosis,evaluation,therapy monitoring,and follow-up.A qualitative medical image is transformed into a quantitative biomarker or surrogate endpoint useful in drug design decision-making.For this,a parameter needs to be identified that characterizes the disease baseline and its subsequent response to treatment.The result is a quantifiable improvement in healthcare quality in most therapeutic areas,resulting in improvements in quality and life duration.This paper provides an overview of recent studies on applying the deep learning method in medical imaging and drug design.We briefly discuss the fields related to the history of deep learning,medical imaging,and drug design.

Effect of Crossbreeding on Beef Production of Jersey Herd Using Fleckvieh Sires Maintained on a Pasture-Based Feeding System

Jersey bull calves and steers produce high quality meat;however, the growth rate of Jersey bull calves for veal and beef are low in comparison to other dairy breeds. This could be improved by crossbreeding with beef breeds. In the current study, the veal and beef production of Jersey and Fleckvieh × Jersey (F × J) bull calves and steers were compared. Bull calves were reared intensively for veal up to a carcass weight not exceeding 100 kg, or as steers for beef to 21 months of age. In both veal and steer production systems, the mean birth weight of F × J bull calves was higher (P < 0.001) than Jersey bull calves, i.e. 33.5 ± 1.2 kg vs. 27.9 ± 1.2 kg and (33.4 ± 0.9 kg vs. 26.9 ± 0.9 kg, respectively. The body weight of Jersey and F × J veal calves at 6 months of age differed (P < 0.01) being 163.5 ± 3.9 and 180.6 ± 4.0 kg respectively. This could be attributed to a higher (P < 0.01) average daily gain (ADG) of 0.82 ± 0.02 kg/day for F × J compared to 0.73 ± 0.02 kg/day for Jersey bull calves. Marketing age in the veal production system differed (P < 0.001) with Jersey and F×J bull calves marketed at 7.1 ± 0.1 and 6.3 ± 0.1 months of age, respectively. End live weight at 21 months of age of Jersey and F × J differed (P < 0.01) being 322.6 ± 13.4 and 441.4 ± 14.9 kg respectively. This was due to a higher (P < 0.01) average daily gain in F × J vs. Jersey steers, i.e. 0.64 ± 0.02 and 0.46 ± 0.0 kg/day, respectively. Fleckvieh × Jersey steers had a higher (P < 0.01) carcass weight at 21 months of age being 206.5 ± 8.9 kg vs. 157.9 ± 8.6 kg respectively. These results indicated the potential of improving beef production characteristics of the Jersey cattle through crossbreeding.

The genomics of desmoplastic small round cell tumor reveals the deregulation of genes related to DNA damage response, epithelial-mesenchymal transition, and immune response

Background:Desmoplastic small round cell tumor(DSRCT)is a rare,aggressive,and poorly investigated simple sarcoma with a low frequency of genetic deregulation other than an Ewing sarcoma RNA binding protein 1(EWSR1)-Wilm’s tumor suppressor(WT1)translocation.We used whole-exome sequencing to interrogate six consecutive pretreated DSRCTs whose gene expression was previously investigated.Methods:DNA libraries were prepared from formalin-fixed,paraffin-embedded archival tissue specimens following the Agilent SureSelectXT2 target enrichment protocol and sequenced on Illumina NextSeq 500.Raw sequence data were aligned to the reference genome with Burrows-Wheeler Aligner algorithm.Somatic mutations and copy number alterations(CNAs)were identified using MuTect2 and EXCAVATOR2,respectively.Biological functions associated with altered genes were investigated through Ingenuity Pathway Analysis(IPA)software.Results:A total of 137 unique somatic mutations were identified:133 mutated genes were case-specific,and 2 were mutated in two cases but in different positions.Among the 135 mutated genes,27%were related to two biological categories:DNA damage-response(DDR)network that was also identified through IPA and mesenchymal-epithelial reverse transition(MErT)/epithelial-mesenchymal transition(EMT)already demonstrated to be relevant in DSRCT.The mutated genes in the DDR network were involved in various steps of transcription and particularly affected pre-mRNA.Half of these genes encoded RNA-binding proteins or DNA/RNA-binding proteins,which were recently rec-ognized as a new class of DDR players.CNAs in genes/gene families,involved in MErT/EMT and DDR,were recurrent across patients and mostly segregated in the MErT/EMT category.In addition,recurrent gains of regions in chromosome 1 involving many MErT/EMT gene families and loss of one arm or the entire chromosome 6 affecting relevant immune-regulatory genes were recorded.Conclusions:The emerging picture is an extreme inter-tumor heterogeneity,characterized by the concurrent deregulation of the DDR and MErT/EMT dynamic and plastic programs that could favour genomic instability and explain the refractory DSRCT profile.

Local grid refinement in multigrid method for point contact problems of polycrystalline anisotropic material under dry and lubricated conditions

Predicting rolling bearing fatigue life requires knowledge of the three-dimensional(3D)stress fields in the roller and raceway near the lubricated contact.Owing to the increasingly severe operating conditions,the effect of localized features such as surface roughness,subsurface inclusions,and even the crystallographic structure of the material becomes important.Achieving such detail requires(locally)extremely dense gridding in simulations,which in 3D is a major challenge.Multigrid techniques have been demonstrated to be capable of solving such problems.In this study,multigrid techniques are shown to further increase the efficiency of the solution by exploiting local grid refinement while maintaining the simplicity of a uniform discretization.This is achieved by employing increasingly finer grids only locally,where the highest resolution is required.Results are presented for dry contact and elastohydrodynamically lubricated contact cases,circular as well as elliptic,with varying crystallographic structure,and with surface roughness.The results show that the developed algorithm is very well suited for detailed analysis,with also excellent prospects for computational diagnostics involving actual material crystallographic structure from electron backscatter diffraction measurements.

Gestational weight gain among minority adolescents predicts term birth weight

Background In adolescents, there is limited evidence on the independent and additive effect of prepregnancy body mass index (BMI) and gestational weight gain on infant birth weight. Data also show that this effect may vary by race. We sought to examine the impact of maternal prepregnancy BMI and gestational weight gain on birth weight and risk of large for ges-tational age (LGA) in term newborns of minority adolescent mothers. Methods This was a retrospective cohort study of 411 singleton live term infants born to mothers ≤ 18 years. Data were abstracted from electronic medical records. Results Gestational weight gain was related to infant birth weight (ρ = 0.36,P < 0.0001), but BMI was not (ρ = 0.025, P = 0.61). On regression analysis, gestational weight gain, gestational age and Hispanic ethnicity were independent predic-tors of birth weight, controlling for maternal age, BMI, parity, tobacco/drug use and preeclampsia. The probability of having an LGA infant increased with weight gain [adjusted odds ratio (aOR) 1.14, 95% confidence interval (CI) 1.07–1.21] but not with BMI. Mothers who gained weight in excess of 2009 Institute of Medicine (IOM) recommendations had a greater risk of having an LGA infant compared to those who gained within recommendations (aOR 5.7, 95% CI 1.6–19.5). Conclusions Minority adolescents with greater gestational weight gain had infants with higher birth weight and greater risk of LGA;BMI was not associated with either outcome. Further studies are needed to examine the applicability of the 2009 BMI-specific IOM gestational weight gain recommendations to adolescents in minority populations.