Power Transformer Fault Diagnosis Using Random Forest and Optimized Kernel Extreme Learning Machine

Power transformer is one of the most crucial devices in power grid.It is significant to determine incipient faults of power transformers fast and accurately.Input features play critical roles in fault diagnosis accuracy.In order to further improve the fault diagnosis performance of power trans-formers,a random forest feature selection method coupled with optimized kernel extreme learning machine is presented in this study.Firstly,the random forest feature selection approach is adopted to rank 42 related input features derived from gas concentration,gas ratio and energy-weighted dissolved gas analysis.Afterwards,a kernel extreme learning machine tuned by the Aquila optimization algorithm is implemented to adjust crucial parameters and select the optimal feature subsets.The diagnosis accuracy is used to assess the fault diagnosis capability of concerned feature subsets.Finally,the optimal feature subsets are applied to establish fault diagnosis model.According to the experimental results based on two public datasets and comparison with 5 conventional approaches,it can be seen that the average accuracy of the pro-posed method is up to 94.5%,which is superior to that of other conventional approaches.Fault diagnosis performances verify that the optimum feature subset obtained by the presented method can dramatically improve power transformers fault diagnosis accuracy.

Bioinspired nanomaterials for the treatment of bacterial infections

Infectious diseases pose a serious threat to global health.Although immunizations can control most viral infections,bacterial infections,particularly those caused by drug-resistant strains,continue to cause high rates of illness and death.Unfortunately,the creation of new antibiotics has come to a grinding halt in the last ten years.In response to this crisis,nanotechnology has emerged as a hopeful solution to tackle drug resistance and enhance treatment results.A large variety of biomimetic nanomaterials,termed nanozymes,have demonstrated strong antimicrobial efficacy.While the inherent toxicity of nanomaterials is a concern,recent studies have harnessed the stimuli-responsiveness of nanomaterials to enable local and/or targeted delivery to reduce the treatment side effects.Indeed,the physicochemical versatility of nanomaterials affords many degrees of freedom that enable rational design of smart or autonomous therapeutics,which cannot be achieved using conventional antibiotics.The design straddles the fields of catalysis,nanoscience,microbiology,and translational medicine.To provide an overview of this interdisciplinary landscape,this review is organized based on composition into lipid,metal,metal oxide,and non-metallic nanomaterials.Liposomes as a delivery vehicle enhance bioavailability and reduce toxicity.Metal-and metal oxide-based nanomaterials inhibit bacterial growth by mimicking natural enzymatic activities such as peroxidase(POD)and oxidase.Furthermore,carbon-,polymer-,and cell membrane-based nanomaterials are combined into a discussion on non-metallic materials.At the end of this review,potentially fruitful directions for future bioinspired nanomaterials in infectious disease treatment are included.

贝叶斯推理和动态神经反馈促进先天性心脏病智能诊断的临床应用

先天性心脏病(CHD)是婴幼儿死亡主要原因。基于人工智能的先天性心脏病诊断网络(CHDNet)是一种基于超声心动图视频的二分类模型,用于判别超声心动图视频是否包含心脏缺陷。现有的CHDNet模型表现出与医学专家相当甚至更好的判别性能,但它们在训练集之外的样本上的不可靠性已成为模型部署的关键瓶颈。而这是当前大多数基于AI诊断方法的共性问题。为了克服这一挑战,本文我们提出了两种基本机制——贝叶斯推理和动态神经反馈——分别用于衡量和提高人工智能诊断的可靠性。贝叶斯推理允许神经网络模型输出CHD判别的可靠性而不仅仅是单一的判别结果,而动态神经反馈是一个计算神经反馈单元,允许神经网络将知识从输出层反馈给浅层,使神经网络有选择地激活相关神经元。为了评估这两种机制的有效性,我们在包含三种常见CHD缺陷的4151个超声心动图视频上训练了CHDNet,并在1037个超声心动图视频的内部测试集和从其他心血管成像设备新收集的692个外部视频集上对其进行了测试。每个超声心动图视频对应于一位患者和一次就诊。我们在多种代表性神经网络架构上展示了贝叶斯推理获得的可靠性如何解释和量化神经网络内部与外部测试集之间的性能显著差异,以及尽管输入被噪声破坏或使用外部测试集时,设计的反馈单元如何帮助神经网络保持高精度和可靠性。

Fault Diagnosis of Power Transformer Based on Improved ACGAN Under Imbalanced Data

The imbalance of dissolved gas analysis(DGA)data will lead to over-fitting,weak generalization and poor recognition performance for fault diagnosis models based on deep learning.To handle this problem,a novel transformer fault diagnosis method based on improved auxiliary classifier generative adversarial network(ACGAN)under imbalanced data is proposed in this paper,which meets both the requirements of balancing DGA data and supplying accurate diagnosis results.The generator combines one-dimensional convolutional neural networks(1D-CNN)and long short-term memories(LSTM),which can deeply extract the features from DGA samples and be greatly beneficial to ACGAN’s data balancing and fault diagnosis.The discriminator adopts multilayer perceptron networks(MLP),which prevents the discriminator from losing important features of DGA data when the network is too complex and the number of layers is too large.The experimental results suggest that the presented approach can effectively improve the adverse effects of DGA data imbalance on the deep learning models,enhance fault diagnosis performance and supply desirable diagnosis accuracy up to 99.46%.Furthermore,the comparison results indicate the fault diagnosis performance of the proposed approach is superior to that of other conventional methods.Therefore,the method presented in this study has excellent and reliable fault diagnosis performance for various unbalanced datasets.In addition,the proposed approach can also solve the problems of insufficient and imbalanced fault data in other practical application fields.

Research Progress on the Development and Application of New Natural Tobacco Humectants

The research progress on the development and application of new natural tobacco humectants was reviewed in the paper.Plants,animals and microbes contain special moisturizing substances,showing distinctive advantages and outstanding effects in respect of the tobacco moisture content maintaining and cigarette suction comfort promoting.The development necessity,major types,functional mechanism and influence on cigarette quality of new natural tobacco humectants development were compared and elaborated.Additionally,the prospect of new natural tobacco humectants development and application was forecasted.

Realizing n-type CdSb with promising thermoelectric performance

Realizing high performance in both n-type and p-type materials is essential for designing efficient ther-moelectric devices.However,the doping bottleneck is often encountered,i.e.,only one type of conduction can be realized.As one example,p-type CdSb with high thermoelectric performance has been discovered for several decades,while its n-type counterpart has rarely been reported.In this work,the calculated band structure of CdSb demonstrates that the valley degeneracy is as large as ten for the conduction band,and it is only two for the valence band.Therefore,the n-type CdSb can potentially realize an ex-ceptional thermoelectric performance.Experimentally,the n-type conduction has been successfully real-ized by tuning the stoichiometry of CdSb.By further doping indium at the Cd site,an improved room-temperature electron concentration has been achieved.Band modeling predicts an optimal electron con-centration of∼2.0×1019 cm−3,which is higher than the current experimental values.Therefore,future optimization of the n-type CdSb should mainly focus on identifying practical approaches to optimize the electron concentration.

Transcript profile of CLSTN3B gene in human white adipose tissue is associated with obesity and mitochondrial gene program

Dear Editor,Obesity is a major health issue with global prevalence and increases the risk of many metabolic diseases.Of particular con-cern is the increasing incidence of type 2 diabetes,the primary causes of which are obesity-driven insulin resistance in white adipose tissue(WAT),skeletal muscle,and liver,and decreased insulin secretion by pancreaticβ-cells[1].

An Improved SVM Based Wind Turbine Multi-fault Detection Method

A fault detection method bases on wind turbines(WTs)supervisory control and data acquisition(SCADA)is proposed,principal component analysis(PCA)was used to reduce the dimension of target features to 1-D,so that PCA output 1-D data can be used as label of support vector machine(SVM).Thus on the premise of not losing the prediction correctness,one model can detect the fault of 2 to 4 features,largely reduce the complexity of model building.Different experiments are present to show the effectiveness of the proposed method.

Ensemble Learning-Based Wind Turbine Fault Prediction Method with Adaptive Feature Selection

In this paper we present a wind turbine (WT) fault detection method based on ensemble learning, WT supervisory control and data acquisition (SCADA) is used for model building. In feature selection process, random forest algorithm is applied to get the feature importances,this is much convenient compared with general feature selection by experience, also more accurate result is obtain. In model building,SVM based bagging algorithm is used, compared to individual SVM,out method is much faster and again with a better result.

VaR Criteria for optimal limited changeloss and truncated change-loss reinsurance

Reinsurance can provide an effective way for insurer to manage its risk exposure. In this paper, we further analyze the optimal reinsurance models recently proposed by J. Cai and K. S. Tan [Astin Bulletin, 2007, 37（1）： 93- 112]. With the criteria of minimizing the value-at-risk （VaR） risk measure of insurer＇s total loss exposure, we derive the optimal values of sharing proportion a, retention d, and layer 1 of two reinsurance treaties： the limited change- loss f（x） = a{（x - d）＋ - （x -l）＋} and the truncated change-loss f（x） = a（x - d）＋I（x≤t）. Both of the reinsurance plans have been considered to be more realistic and practical in the real business. Our solutions have several appealing features： （i） there is only one condition to verify for the existence of optimal limited change-loss reinsurance while there always exists an optimal truncated change-loss reinsurance, （ii） the resulting optimal parameters have simple analytic forms which depend only on assumed loss distribution, reinsurer＇s safety loading, and insurer＇s risk tolerance, （iii） the optimal retention d for limited change-loss reinsurance is the same as that by Cai and Tan while the optimal value is smaller for truncated change-loss, （iv） the optimal sharing proportion and layer are always the same for both reinsurance plans, （v） minimized VaR are strictly lower than the values derived by Cai and Tan, （vi） the optimization results reveal possible drawbacks of VaR-based risk management that a heavy tail risk exposure may be expressed by lower VaR.

Decreasing complexity of glucose time series derived from continuous glucose monitoring is correlated with deteriorating glucose regulation

Most information used to evaluate diabetic statuses is collected at a special time-point,such as taking fasting plasma glucose test and providing a limited view of individual’s health and disease risk.As a new parameter for continuously evaluating personal clinical statuses,the newly developed technique“continuous glucose monitoring”(CGM)can characterize glucose dynamics.By calculating the complexity of glucose time series index(CGI)with refined composite multi-scale entropy analysis of the CGM data,the study showed for the first time that the complexity of glucose time series in subjects decreased gradually from normal glucose tolerance to impaired glucose regulation and then to type 2 diabetes(P for trend<0.01).Furthermore,CGI was significantly associated with various parameters such as insulin sensitivity/secretion(all P<0.01),and multiple linear stepwise regression showed that the disposition index,which reflectsβ-cell function after adjusting for insulin sensitivity,was the only independent factor correlated with CGI(P<0.01).Our findings indicate that the CGI derived from the CGM data may serve as a novel marker to evaluate glucose homeostasis.

秀丽隐杆线虫RNA结合蛋白复合物AMG-1/SLRP-1通过线粒体稳态维持实现调控生殖腺发育和精子发生

RNA结合蛋白(mtRBP)介导的mRNA转录后调节对精子发生必不可少但却鲜有报道.在本文中,我们鉴定到一个在生殖腺中特异性表达的线粒体RNA结合蛋白AMG-1,它是秀丽隐杆线虫精子发生过程中必需的蛋白,同时与哺乳动物LRPPRC蛋白同源.amg-1突变会阻碍生殖腺的发育,最终导致生殖细胞的线粒体形态和结构异常以及线粒体功能障碍.通过测序鉴定RNA结合蛋白的靶点发现,AMG-1更倾向于与mtDNA编码的参与线粒体核糖体组装的12S和16S核糖体RNA(rRNA)结合,12S rRNA对于维持生殖细胞线粒体蛋白稳态至关重要,而12S rRNA的表达却受AMG-1蛋白调节.此外,哺乳动物SLIRP在秀丽线虫中的同源蛋白SLRP-1蛋白与AMG-1在遗传上存在互作关系,它们可共同调节秀丽线虫的精子发生和育性.综上所述,这些发现揭示了mtRBP蛋白AMG-1在线粒体调控中的新机制,这可能为由线粒体功能障碍引发的男性不育治疗提供新的理论基础.

Research progress in the application of time in range: more than a percentage

Glucose monitoring is an important part of medical care in diabetes mellitus,which not only helps assess glycemic control and treatment safety,but also assists with treatment adjustment.With the development of continuous glucose monitoring(CGM),the use of CGM has increased rapidly.With the wealth of glucose data produced by CGM,new metrics are greatly needed to optimally evaluate glucose status and guide the treatment.One of the parameters that CGM provides,time in range(TIR),has been recognized as a key metric by the international consensus.Before the adoption of TIR in clinical practice,several issues including the minimum length of CGM use,the setting of the target range,and individualized TIR goals are summarized.Additionally,we discussed the mounting evidence supporting the association between TIR and diabetes-related outcomes.As a novel glucose metric,it is of interest to compare TIR with other conventional glucose markers such as glycated hemoglobin A1c.It is anticipated that the use of TIR may provide further information on the quality of glucose control and lead to improved diabetes management.

：CL5 as a potential immunotherapeutic target in triple-negative breast cancer

Breast cancer （BC） is a leading cause of mortality among women in the world. To date, a number of molecules have been established as disease status indicators and therapeutic targets. The best known among them are estrogen receptor-α （ER-α）, progesterone receptor （PR） and H ER-2/neu. About 15%-20% BC patients do not respond effectively to thera pies targeting these classes of tumor-promoting factors. Thus, additional targets are strongly and urgently sought after in therapy for human BCs negative for ER, PR and HER-2, the so-called triple-negative BC （TNBC）. Recent clinical work has revealed that CC chemokine ligand 5 （CCL5） is strongly associated with the progression of BC, particularly TNBC. How CCL5 contributes to the development of TN BC is not well understood. Experimental animal studies have begun to address the mechanistic issue. In this article, we will review the clinical and laboratory work in this area that has led to our own hypothesis that targeting CCL5 in TNBCs will have favorable therapeutic outcomes with minimal adverse impact on the general physiology.

Impact of clinical research on public health policy of neonatal screening for congenital heart disease in China

Congenital heart disease(CHD)is one of the most common congenital malformations,with a prevalence of approximately 1%in live births.[1,2]Among them,1/4 to 1/3 are major CHD,including critical CHD(CCHD)and serious CHD,which require surgical or catheter intervention before 1 year of age.[3]Major CHD,if left untreated,may cause serious outcomes,such as heart failure,cardiogenic shock,acidosis,and hypoxic-ischemic brain damage,and is one of the leading causes of infant death and childhood disability.[4]The missed diagnosis rate of CCHD in four nurseries ranges from 13%to 48%in developed countries,[4,5]but in China,is>71%before serious symptoms occur after birth.[6]Therefore,screening for CHD is useful for early detection,diagnosis,and treatment and helps to improve prognosis and reduces the medical,economic,and mental burden to both the family and the society.[5,7]

Insulin sensitivity,βcell function,and adverse pregnancy outcomes in women with gestational diabetes

Background: The potential impact of β cell function and insulin sensitivity on adverse pregnancy outcomes in women with gestational diabetes mellitus (GDM) remains uncertain. We aimed to investigate the association between β cell dysfunction, insulin resistance, and the composite adverse pregnancy outcomes.Methods: This observational study included 482 women diagnosed with GDM during pregnancy. Quantitative metrics on β cell function and insulin sensitivity during pregnancy were calculated using traditional equations. The association of β cell dysfunction and insulin resistance with the risk of the composite adverse pregnancy outcomes was investigated using multivariable-adjusted logistic regression models.Results: Multivariable-adjusted odds ratios (ORs) of adverse pregnancy outcomes across quartiles of homeostatic model assessment for insulin resistance (HOMA-IR) were 1.00, 0.95, 1.34, and 2.25, respectively (P for trend = 0.011). When HOMA-IR was considered as a continuous variable, the multivariable-adjusted OR of adverse pregnancy outcomes was 1.34 (95% confidence interval 1.16-1.56) for each 1-unit increase in HOMA-IR. Multivariable-adjusted ORs of adverse pregnancy outcomes across quartiles of homeostatic model assessment for β cell function (HOMA-β) were 1.00, 0.51, 0.60, and 0.53, respectively (P for trend = 0.068). When HOMA-β was considered as a continuous variable, the multivariable-adjusted OR of adverse pregnancy outcomes was 0.57 (95% CI 0.24-0.90) for each 1-unit increase in HOMA-β. However, other quantitative metrics were not associated with the composite adverse pregnancy outcomes.Conclusions: We demonstrated a significant association of β cell function and insulin sensitivity with the risk of adverse pregnancy outcomes. We have provided additional evidence on the early identification of adverse pregnancy outcomes besides the glycemic values.

Establishment of Novel Monoclonal Fabs Specific for Epstein-Barr Virus Encoded Latent Membrane Protein 1

Dear Editor,Epstein-Barr virus(EBV,also termed human herpesvirus-4)was the first identified human tumor virus.Since its discovery in 1964,studies have shown that EBV infects over 90%of all people by the time they are adults(Williams and Crawford 2006).EBV infection can result in mucocutaneous and systemic diseases,ranging from selflimited illnesses to aggressive malignancies,including B cell Hodgkin lymphoma and nasopharyngeal carcinoma.In vitro,EBV transforms resting B cells into proliferating blast cells(Pope et al.1968).

Long non-coding RNA SAP30-2:1 is downregulated in congenital heart disease and regulates cell proliferation by targeting HAND2

Congenital heart disease(CHD)is the most common birth defect worldwide.Long non-coding RNAs(lncRNAs)have been implicated in many diseases.However,their involvement in CHD is not well understood.This study aimed to investigate the role of dysregulated lncRNAs in CHD.We used Gene Expression Omnibus data mining,bioinformatics analysis,and analysis of clinical tissue samples and observed that the novel lncRNA SAP30-2:1 with unknown function was significantly downregulated in damaged cardiac tissues from patients with CHD.Knockdown of lncRNA SAP30-2:1 inhibited the proliferation of human embryonic kidney and AC16 cells and decreased the expression of heart and neural crest derivatives expressed 2(HAND2).Moreover,lncRNA SAP30-2:1 was associated with HAND2 by RNA immunoprecipitation.Overall,these results suggest that lncRNA SAP30-2:1 may be involved in heart development through affecting cell proliferation via targeting HAND2 and may thus represent a novel therapeutic target for CHD.

Low-carbohydrate diets lead to greater weight loss and better glucose homeostasis than exercise: a randomized clinical trial

Lifestyle interventions,including dietary adjustments and exercise,are important for obesity management.This study enrolled adults with overweight or obesity to explore whether either low-carbohydrate diet(LCD)or exercise is more effective in metabolism improvement.Forty-five eligible subjects were randomly divided into an LCD group(n=22)and an exercise group(EX,n=23).The subjects either adopted LCD(carbohydrate intake<50 g/day)or performed moderate-to-vigorous exercise(≥30 min/day)for 3 weeks.After the interventions,LCD led to a larger weight loss than EX(−3.56±0.37 kg vs.−1.24±0.39 kg,P<0.001),as well as a larger reduction in fat mass(−2.10±0.18 kg vs.−1.25±0.24 kg,P=0.007)and waist circumference(−5.25±0.52 cm vs.−3.45±0.38 cm,P=0.008).Both interventions reduced visceral and subcutaneous fat and improved liver steatosis and insulin resistance.Triglycerides decreased in both two groups,whereas low-density lipoprotein cholesterol increased in the LCD group but decreased in the EX group.Various glycemic parameters,including serum glycated albumin,mean sensor glucose,coefficient of variability(CV),and largest amplitude of glycemic excursions,substantially declined in the LCD group.Only CV slightly decreased after exercise.This pilot study suggested that the effects of LCD and exercise are similar in alleviating liver steatosis and insulin resistance.Compared with exercise,LCD might be more efficient for weight loss and glucose homeostasis in people with obesity.

超声心动图在肥厚型心肌病诊断中的应用进展

肥厚型心肌病(hypertrophic cardiomyopathy,HCM)是一类肌小节蛋白编码基因变异、以左心室心肌肥厚为主要特征的心脏疾病。HCM具有明显的家族史,遗传风险为50%[1]。