Ice loads on a ship hull affect the safety of the hull structure and the ship maneuvering performance in ice-covered regions. A discrete element method （DEM） is used to simulate the interaction between drifting ice ...Ice loads on a ship hull affect the safety of the hull structure and the ship maneuvering performance in ice-covered regions. A discrete element method （DEM） is used to simulate the interaction between drifting ice floes and a moving ship. The pancake ice floes are modelled with three-dimensional （3-D） dilated disk elements considering the buoyancy, drag force and additional mass induced by the current. The ship hull is modelled with 3D disks with overlaps. Ice loads on the ship hull are determined through the contact detection between ice floe element and ship hull element and the contact force calculation. The influences of different ice conditions （current velocities and directions, ice thicknesses, concentrations and ice floe sizes） and ship speeds are also examined on the dynamic ice force. The simulated results are compared qualitatively well with the existing field data and other numerical results. This work can be helpful in the shil3 structure design and the navigation securitv in ice-covered fields.展开更多
目的:评价基于深度学习的继发性肺结核CT辅助诊断模型在临床应用中的价值。方法:回顾性收集2018年12月至2023年4月在重庆市公共卫生医疗救治中心接受胸部CT平扫的2004例患者的病例资料,分为肺部正常组(544例)、普通肺部感染组(526组)和...目的:评价基于深度学习的继发性肺结核CT辅助诊断模型在临床应用中的价值。方法:回顾性收集2018年12月至2023年4月在重庆市公共卫生医疗救治中心接受胸部CT平扫的2004例患者的病例资料,分为肺部正常组(544例)、普通肺部感染组(526组)和继发性肺结核组(934例)。按照随机分组(通过R语言的sample函数实现训练集和测试集的完全随机分组)的方式,将数据集划分为训练集(1402例,70.0%)和测试集(602例,30.0%)。所有图像采用肺野自动分割算法,获得肺野区域。进一步采用BasicNet和DenseNet算法进行三组间的分类研究。采用曲线下面积(area under curve,AUC)、敏感度、特异度和准确率评价模型的分类性能。最后,在测试数据中,将最优模型与3位不同年资的放射科医生的诊断结果进行比较。结果:602例独立测试集中,DenseNet模型的性能优于BasicNet模型,两种模型的平均AUC、敏感度、特异度和准确率分别为92.1%和89.4%、79.7%和74.0%、89.4%和86.6%、86.2%和83.3%。其中,DenseNet模型的诊断性能优于低年资医生(准确率分别为90.7%和89.1%,Kappa=0.677),与中年资和高年资医生的诊断水平(准确率分别为90.7%、92.2%和95.3%,Kappa值分别为0.746和0.819)保持高度一致性。结论:DenseNet模型能较准确地识别继发性肺结核,与放射科中年资医师的诊断水准相当,可以作为继发性肺结核的辅助诊断工具。展开更多
Objective:To improve the diagnosis and therapeutic effect of occurrence and development of hyponatremia and disorder of acid-base balance among patients with hepatic encephalopathy(HE) by elucidating the regularity an...Objective:To improve the diagnosis and therapeutic effect of occurrence and development of hyponatremia and disorder of acid-base balance among patients with hepatic encephalopathy(HE) by elucidating the regularity and mechanism,as well as its influence on prognosis.Methods:327 HE patients admitted to our hospital from January 1990 to June 2010 were enrolled.Meanwhile 316 patients hospitalized in the medical department of the same hospital were chosen as the control group.Patients in both groups were given the same methods to measure arterial blood gas parameters(pH value,PaCO2,[HCO3-],TCO2,BE and SaO2),blood biochemistry([Na+],[K+],[Cl-]),liver function,kidney function and blood glucose,serum sodium,and thereupon tocalculate the anion gap(AG) and the potential [HCO3-],and acid-base balance disorder.Results:Among the 327 HE patients,hyponatremia was found in 188 cases(57.4%),of whom 132 patients died(70.2%).While among the 316 patients in control group,68 presented with hyponatremia(21.5%),and 19 died(27.9%).The incidence and mortality were significantly different between the two groups(P<0.001).All the 327 patients presented with different degrees of acid-base balance disorder and 178 died(54.4%),in whom 164(50.2%) belonged to simple acid-base balance disorder and 74(45.1%) died,136(41.6%) were dual acid-base balance disorder and 80(58.8%) died,27(8.2%) were triple acid-base disturbance and 24(88.9%) died.Whereas in the control group only 83 patients(26.2%) were recognized as simple and dual acid-base balance disorder,and 18(21.7%) died.There was higher incidence of acid-base balance disorder and mortality rate in HE group than control one(P<0.001).Conclusion:Hyponatremia is valuable to judge HE patients' prognosis.The key parameters in the judgment and evaluation on acid-base balance disorder among HE patients are the change of pH values and serum electrolyte values.When pH value ≤ 7.30 or > 7.55,it generally suggests a poor prognosis.展开更多
Sea surface winds from reanalysis （NCEP-2 and ERA-40 datasets） and satellite-based products （QuikSCAT and NCDC blended sea winds） are evaluated using in situ ship measurements from the Chinese National Antarctic R...Sea surface winds from reanalysis （NCEP-2 and ERA-40 datasets） and satellite-based products （QuikSCAT and NCDC blended sea winds） are evaluated using in situ ship measurements from the Chinese National Antarctic Research Expeditions （CH1NAREs） from 1989 through 2006, with emphasis on the Southern Ocean （south of 45°S）. Compared with ship observations, the reanalysis winds have a positive mean bias （0.32 m·s-1 for NCEP-2 and 0.13 m·s-1 for ERA-40）, and this bias is more pronounced in the Southern Ocean （0.57 m·s-1 and 0.45 m·s-1, respectively）. However, mean biases are negative in the tropics and subtropics. The satellite-based winds also show positive mean biases, larger than those of the reanalysis data. All four wind products overestimate ship wind speed for weak winds （〈4 m·s-1） but underestimate for strong winds （〉10 m·s-1）. Differences between the reanalysis and satellite winds are examined to identify regions with large discrepancies.展开更多
为探究中国农业碳排放的时空分布特征及驱动因素,基于2000—2021年中国31个省(自治区、直辖市,不包括香港地区、澳门地区和台湾地区,下同)统计年鉴数据,考察水利用、土地利用和能源消耗的碳排放,利用联合国政府间气候变化专门委员会(IP...为探究中国农业碳排放的时空分布特征及驱动因素,基于2000—2021年中国31个省(自治区、直辖市,不包括香港地区、澳门地区和台湾地区,下同)统计年鉴数据,考察水利用、土地利用和能源消耗的碳排放,利用联合国政府间气候变化专门委员会(IPCC)碳排放因子建立2000—2021年水、土地和能源3个子系统相关变量,计算各省(自治区、直辖市)农业年碳排放总量,结合莫兰指数对农业碳排放时空演变趋势及空间关联特征进行分析,并运用对数均值迪氏分解法(Logarithmic Mean Divisia Index,简称LMDI)探析农业碳排放的主要驱动因素。结果表明:1)从时序变化看,农业碳排放量整体呈倒“V”型变化趋势。2)从农业碳排放来源看,农业源碳排放中源于化肥的碳排放占比最高。3)从农业碳排放空间差异看,碳排放较大的省份(自治区、直辖市)主要集中在黄淮海区域以及中部平原等水土资源条件丰富且优质的地区,西部地区与部分直辖市(北京、上海、天津)农业碳排放量较少,高农业碳排放地区存在向北蔓延的趋势。4)农业碳排放在空间上具有集聚效应,且随着时间推移,集聚效应的显著性有所下降,其中河南、安徽、山东等省份(自治区、直辖市)具有显著的“高-高集聚”效应,北京、天津、青海等省份(自治区、直辖市)具有显著的“低-低集聚”效应。5)农业水资源经济产出因素和农业劳动力密集度因素为正向驱动因素,农业水资源经济产出因素为中国农业碳排放增加的最主要因素;农业生产效率因素、劳动力规模因素和农业水土匹配度因素为碳排放负向驱动因素,其中农业生产效率因素的碳减排贡献率最高,为中国农业碳排放减少的最主要驱动因素。基于以上结果,本文针对中国农业碳减排提出以下建议:政府应加大对低碳农业的投入,支持新型肥料和新能源农机的研发,提高水土资源利用效率。同时,要利用农业碳排放的集聚效应,推动农业集中发展和区域间合作,培养新型农业人才。展开更多
微服务架构因具有良好的可扩展性和可维护性越来越受到云应用软件的青睐.与此同时,微服务之间复杂的交互使得系统的性能异常检测变得更加困难.现有的微服务性能异常检测方法均不能很好地建立跨不同调用路径的微服务及其对应的响应时间...微服务架构因具有良好的可扩展性和可维护性越来越受到云应用软件的青睐.与此同时,微服务之间复杂的交互使得系统的性能异常检测变得更加困难.现有的微服务性能异常检测方法均不能很好地建立跨不同调用路径的微服务及其对应的响应时间之间的复杂关系,导致异常检测准确率不高、根因定位不准确.提出了一种基于Transformer的微服务性能异常检测与根因定位方法TTEDA(Transformer trace explore data analysis).首先将调用链构建为微服务调用序列和对应的响应时间序列,然后借助自注意力机制捕捉微服务之间的调用关系,并通过编码器-解码器建立微服务的响应时间与其调用路径之间的关联关系,从而获得微服务在不同的调用链上的正常响应时间分布.基于学习到的正常模式判断调用链的异常,并可将异常精确到微服务级别.进一步地,利用微服务之间的调用关系以及异常的传播方式,对出现性能异常的微服务进行反向拓扑排序,实现了准确快速的根因定位.在开源基准微服务系统Train-Ticket的数据集和AIops挑战赛数据集评估了TTEDA的有效性,相比于同类异常检测方法AEVB,Multi-LSTM,TraceAnomaly,精确率平均提高了48.6%,30.2%,3.5%,召回率平均提高了34.7%,1.1%,4.1%.相比于根因定位算法MonitorRank和TraceAnomaly,根因定位的准确率分别提高了35.4个百分点和6.1个百分点.展开更多
基金The Special Funding for National Marine Commonwealth Industry of China under contract Nos 201105016 and 2012418007the National Natural Science Foundation of China under contract No.41176012
文摘Ice loads on a ship hull affect the safety of the hull structure and the ship maneuvering performance in ice-covered regions. A discrete element method （DEM） is used to simulate the interaction between drifting ice floes and a moving ship. The pancake ice floes are modelled with three-dimensional （3-D） dilated disk elements considering the buoyancy, drag force and additional mass induced by the current. The ship hull is modelled with 3D disks with overlaps. Ice loads on the ship hull are determined through the contact detection between ice floe element and ship hull element and the contact force calculation. The influences of different ice conditions （current velocities and directions, ice thicknesses, concentrations and ice floe sizes） and ship speeds are also examined on the dynamic ice force. The simulated results are compared qualitatively well with the existing field data and other numerical results. This work can be helpful in the shil3 structure design and the navigation securitv in ice-covered fields.
文摘目的:评价基于深度学习的继发性肺结核CT辅助诊断模型在临床应用中的价值。方法:回顾性收集2018年12月至2023年4月在重庆市公共卫生医疗救治中心接受胸部CT平扫的2004例患者的病例资料,分为肺部正常组(544例)、普通肺部感染组(526组)和继发性肺结核组(934例)。按照随机分组(通过R语言的sample函数实现训练集和测试集的完全随机分组)的方式,将数据集划分为训练集(1402例,70.0%)和测试集(602例,30.0%)。所有图像采用肺野自动分割算法,获得肺野区域。进一步采用BasicNet和DenseNet算法进行三组间的分类研究。采用曲线下面积(area under curve,AUC)、敏感度、特异度和准确率评价模型的分类性能。最后,在测试数据中,将最优模型与3位不同年资的放射科医生的诊断结果进行比较。结果:602例独立测试集中,DenseNet模型的性能优于BasicNet模型,两种模型的平均AUC、敏感度、特异度和准确率分别为92.1%和89.4%、79.7%和74.0%、89.4%和86.6%、86.2%和83.3%。其中,DenseNet模型的诊断性能优于低年资医生(准确率分别为90.7%和89.1%,Kappa=0.677),与中年资和高年资医生的诊断水平(准确率分别为90.7%、92.2%和95.3%,Kappa值分别为0.746和0.819)保持高度一致性。结论:DenseNet模型能较准确地识别继发性肺结核,与放射科中年资医师的诊断水准相当,可以作为继发性肺结核的辅助诊断工具。
文摘Objective:To improve the diagnosis and therapeutic effect of occurrence and development of hyponatremia and disorder of acid-base balance among patients with hepatic encephalopathy(HE) by elucidating the regularity and mechanism,as well as its influence on prognosis.Methods:327 HE patients admitted to our hospital from January 1990 to June 2010 were enrolled.Meanwhile 316 patients hospitalized in the medical department of the same hospital were chosen as the control group.Patients in both groups were given the same methods to measure arterial blood gas parameters(pH value,PaCO2,[HCO3-],TCO2,BE and SaO2),blood biochemistry([Na+],[K+],[Cl-]),liver function,kidney function and blood glucose,serum sodium,and thereupon tocalculate the anion gap(AG) and the potential [HCO3-],and acid-base balance disorder.Results:Among the 327 HE patients,hyponatremia was found in 188 cases(57.4%),of whom 132 patients died(70.2%).While among the 316 patients in control group,68 presented with hyponatremia(21.5%),and 19 died(27.9%).The incidence and mortality were significantly different between the two groups(P<0.001).All the 327 patients presented with different degrees of acid-base balance disorder and 178 died(54.4%),in whom 164(50.2%) belonged to simple acid-base balance disorder and 74(45.1%) died,136(41.6%) were dual acid-base balance disorder and 80(58.8%) died,27(8.2%) were triple acid-base disturbance and 24(88.9%) died.Whereas in the control group only 83 patients(26.2%) were recognized as simple and dual acid-base balance disorder,and 18(21.7%) died.There was higher incidence of acid-base balance disorder and mortality rate in HE group than control one(P<0.001).Conclusion:Hyponatremia is valuable to judge HE patients' prognosis.The key parameters in the judgment and evaluation on acid-base balance disorder among HE patients are the change of pH values and serum electrolyte values.When pH value ≤ 7.30 or > 7.55,it generally suggests a poor prognosis.
基金supported by the National Natural Science Foundation of China(Grant nos.41006115,41076128,41206184)the Marine Science Youth Fund of SOA(Grant no.2010215)the Chinese Polar Environmental Comprehensive Investigation and Assessment Programmes (Grant no.CHINARE2013-04-01).
文摘Sea surface winds from reanalysis （NCEP-2 and ERA-40 datasets） and satellite-based products （QuikSCAT and NCDC blended sea winds） are evaluated using in situ ship measurements from the Chinese National Antarctic Research Expeditions （CH1NAREs） from 1989 through 2006, with emphasis on the Southern Ocean （south of 45°S）. Compared with ship observations, the reanalysis winds have a positive mean bias （0.32 m·s-1 for NCEP-2 and 0.13 m·s-1 for ERA-40）, and this bias is more pronounced in the Southern Ocean （0.57 m·s-1 and 0.45 m·s-1, respectively）. However, mean biases are negative in the tropics and subtropics. The satellite-based winds also show positive mean biases, larger than those of the reanalysis data. All four wind products overestimate ship wind speed for weak winds （〈4 m·s-1） but underestimate for strong winds （〉10 m·s-1）. Differences between the reanalysis and satellite winds are examined to identify regions with large discrepancies.
文摘为探究中国农业碳排放的时空分布特征及驱动因素,基于2000—2021年中国31个省(自治区、直辖市,不包括香港地区、澳门地区和台湾地区,下同)统计年鉴数据,考察水利用、土地利用和能源消耗的碳排放,利用联合国政府间气候变化专门委员会(IPCC)碳排放因子建立2000—2021年水、土地和能源3个子系统相关变量,计算各省(自治区、直辖市)农业年碳排放总量,结合莫兰指数对农业碳排放时空演变趋势及空间关联特征进行分析,并运用对数均值迪氏分解法(Logarithmic Mean Divisia Index,简称LMDI)探析农业碳排放的主要驱动因素。结果表明:1)从时序变化看,农业碳排放量整体呈倒“V”型变化趋势。2)从农业碳排放来源看,农业源碳排放中源于化肥的碳排放占比最高。3)从农业碳排放空间差异看,碳排放较大的省份(自治区、直辖市)主要集中在黄淮海区域以及中部平原等水土资源条件丰富且优质的地区,西部地区与部分直辖市(北京、上海、天津)农业碳排放量较少,高农业碳排放地区存在向北蔓延的趋势。4)农业碳排放在空间上具有集聚效应,且随着时间推移,集聚效应的显著性有所下降,其中河南、安徽、山东等省份(自治区、直辖市)具有显著的“高-高集聚”效应,北京、天津、青海等省份(自治区、直辖市)具有显著的“低-低集聚”效应。5)农业水资源经济产出因素和农业劳动力密集度因素为正向驱动因素,农业水资源经济产出因素为中国农业碳排放增加的最主要因素;农业生产效率因素、劳动力规模因素和农业水土匹配度因素为碳排放负向驱动因素,其中农业生产效率因素的碳减排贡献率最高,为中国农业碳排放减少的最主要驱动因素。基于以上结果,本文针对中国农业碳减排提出以下建议:政府应加大对低碳农业的投入,支持新型肥料和新能源农机的研发,提高水土资源利用效率。同时,要利用农业碳排放的集聚效应,推动农业集中发展和区域间合作,培养新型农业人才。
文摘微服务架构因具有良好的可扩展性和可维护性越来越受到云应用软件的青睐.与此同时,微服务之间复杂的交互使得系统的性能异常检测变得更加困难.现有的微服务性能异常检测方法均不能很好地建立跨不同调用路径的微服务及其对应的响应时间之间的复杂关系,导致异常检测准确率不高、根因定位不准确.提出了一种基于Transformer的微服务性能异常检测与根因定位方法TTEDA(Transformer trace explore data analysis).首先将调用链构建为微服务调用序列和对应的响应时间序列,然后借助自注意力机制捕捉微服务之间的调用关系,并通过编码器-解码器建立微服务的响应时间与其调用路径之间的关联关系,从而获得微服务在不同的调用链上的正常响应时间分布.基于学习到的正常模式判断调用链的异常,并可将异常精确到微服务级别.进一步地,利用微服务之间的调用关系以及异常的传播方式,对出现性能异常的微服务进行反向拓扑排序,实现了准确快速的根因定位.在开源基准微服务系统Train-Ticket的数据集和AIops挑战赛数据集评估了TTEDA的有效性,相比于同类异常检测方法AEVB,Multi-LSTM,TraceAnomaly,精确率平均提高了48.6%,30.2%,3.5%,召回率平均提高了34.7%,1.1%,4.1%.相比于根因定位算法MonitorRank和TraceAnomaly,根因定位的准确率分别提高了35.4个百分点和6.1个百分点.
