期刊文献+

二次检索

题名
关键词
文摘
作者
第一作者
机构
刊名
分类号
参考文献
作者简介
基金资助
栏目信息

年份

学科

期刊

共找到890篇文章
< 1 2 45 >
每页显示 20 50 100
Scientific data products and the data pre-processing subsystem of the Chang'e-3 mission 被引量:1
1
作者 Xu Tan Jian-Jun Liu +7 位作者 Chun-Lai Li Jian-Qing Feng Xin Ren Fen-Fei Wang Wei Yan Wei Zuo Xiao-Qian Wang Zhou-Bin Zhang 《Research in Astronomy and Astrophysics》 SCIE CAS CSCD 2014年第12期1682-1694,共13页
The Chang'e-3 (CE-3) mission is China's first exploration mission on the surface of the Moon that uses a lander and a rover. Eight instruments that form the scientific payloads have the following objectives: (1... The Chang'e-3 (CE-3) mission is China's first exploration mission on the surface of the Moon that uses a lander and a rover. Eight instruments that form the scientific payloads have the following objectives: (1) investigate the morphological features and geological structures at the landing site; (2) integrated in-situ analysis of minerals and chemical compositions; (3) integrated exploration of the structure of the lunar interior; (4) exploration of the lunar-terrestrial space environment, lunar sur- face environment and acquire Moon-based ultraviolet astronomical observations. The Ground Research and Application System (GRAS) is in charge of data acquisition and pre-processing, management of the payload in orbit, and managing the data products and their applications. The Data Pre-processing Subsystem (DPS) is a part of GRAS. The task of DPS is the pre-processing of raw data from the eight instruments that are part of CE-3, including channel processing, unpacking, package sorting, calibration and correction, identification of geographical location, calculation of probe azimuth angle, probe zenith angle, solar azimuth angle, and solar zenith angle and so on, and conducting quality checks. These processes produce Level 0, Level 1 and Level 2 data. The computing platform of this subsystem is comprised of a high-performance computing cluster, including a real-time subsystem used for processing Level 0 data and a post-time subsystem for generating Level 1 and Level 2 data. This paper de- scribes the CE-3 data pre-processing method, the data pre-processing subsystem, data classification, data validity and data products that are used for scientific studies. 展开更多
关键词 Moon: data products -- methods: data pre-processing -- space vehicles:instruments
下载PDF
Automatic Fault Prediction of Wind Turbine Main Bearing Based on SCADA Data and Artificial Neural Network 被引量:2
2
作者 Zhenyou Zhang 《Open Journal of Applied Sciences》 2018年第6期211-225,共15页
As the demand for wind energy continues to grow at exponential rate, reducing operation and maintenance (O & M) costs and improving reliability have become top priorities in wind turbine maintenance strategies. Pr... As the demand for wind energy continues to grow at exponential rate, reducing operation and maintenance (O & M) costs and improving reliability have become top priorities in wind turbine maintenance strategies. Prediction of wind turbine failures before they reach a catastrophic stage is critical to reduce the O & M cost due to unnecessary scheduled maintenance. A SCADA-data based condition monitoring system, which takes advantage of data already collected at the wind turbine controller, is a cost-effective way to monitor wind turbines for early warning of failures. This article proposes a methodology of fault prediction and automatically generating warning and alarm for wind turbine main bearings based on stored SCADA data using Artificial Neural Network (ANN). The ANN model of turbine main bearing normal behavior is established and then the deviation between estimated and actual values of the parameter is calculated. Furthermore, a method has been developed to generate early warning and alarm and avoid false warnings and alarms based on the deviation. In this way, wind farm operators are able to have enough time to plan maintenance, and thus, unanticipated downtime can be avoided and O & M costs can be reduced. 展开更多
关键词 Artificial Neural Network SCADA data Wind TURBINE AUTOMATIC FAULT pre-diction
下载PDF
A study of medication rules of traditional Chinese medicine in the treatment of pre diabetes based on data mining
3
作者 Ji-Ya Sun Qiong-Yang Zhou +4 位作者 Chao Cheng Zhi-Heng Zhao Gui-Tong Zhou Yi-Hua Fan Xin-Ju Li 《Medical Data Mining》 2020年第4期166-176,共11页
Background:To summarize the concerted application and prescription rules of traditional Chinese medicine in the treatment of pre diabetes.Methods:Microsoft Excel 2010 was used to summarize the categories,nature,flavou... Background:To summarize the concerted application and prescription rules of traditional Chinese medicine in the treatment of pre diabetes.Methods:Microsoft Excel 2010 was used to summarize the categories,nature,flavour and channel tropism of drugs.The cluster analysis of high-frequency drugs was carried out by SPSS 22.0,and the association rules of high-frequency drugs were analyzed by Apriori algorithm provided by SPSS modeler 14.0.Results:One hundred and forty-six references were included,including 153 prescriptions and 131 drugs.Their frequency of use is listed in the following order.The top 3 categories of drugs were“Tonifying,Heat-Clearing”,diuresis and“Diffusing Dampness”drugs.The top 5 drugs were Huangqi(Astragali radix),Fuling(Poria),Huanglian(Coptidis rhizoma),Shanyao(Dioscoreae rhizoma),Gegen(Puerariae lobatae radix).The top 3 channel tropism of drugs were spleen,stomach and lung.The top 3 nature of drugs were cold,warm and calm.The top 3 flavour of drugs were sweet,bitter and pungent.The cluster analysis of high-frequency drugs showed that it could be classified into 4 categories:“Benefiting Qi”for promoting production of fluid,“Clearing Heat”and“Eliminating Dampness”,“Nourishing Yin”and“Clearing Heat”,and“Invigorating Spleen”for“Diffusing Dampness”.The results of association rule analysis showed that the combination with the highest degree of confidence and support was Poria-Chenpi(Citri reticulatae pericarpium)-Banxia(Pinelliae rhizoma)-Baizhu(Atractylodis macrocephalae rhizoma)and the combination with the highest frequency was Astragali radix-Puerariae lobatae radix.Conclusion:The pre diabetes is due to deficiency.The disease location is spleen and stomach and the pathological factor is phlegm-damp,that is why benefiting qi and invigorating spleen is regarded as the key link of clinical treatment. 展开更多
关键词 pre diabetes Traditional Chinese medicine Medication rules data mining
下载PDF
基于COM Express的回波预处理模块设计 被引量:2
4
作者 潘奇 倪卫芳 张宏超 《现代雷达》 CSCD 北大核心 2012年第6期30-34,共5页
针对舰载雷达数据处理平台的改造需求,设计了一种基于COM Express的回波预处理模块。通过调整雷达数据处理平台及软件的实现方式,将回波预处理前移,使用最小的改动量解决了雷达系统的回波数据接收瓶颈问题。分析了数据处理平台的改造需... 针对舰载雷达数据处理平台的改造需求,设计了一种基于COM Express的回波预处理模块。通过调整雷达数据处理平台及软件的实现方式,将回波预处理前移,使用最小的改动量解决了雷达系统的回波数据接收瓶颈问题。分析了数据处理平台的改造需求,介绍了模块的设计方案、模块的硬件设计、回波预处理的数据传输流程以及PEX8311桥片在Vx-works下的驱动设计。工程实践表明,该模块大大提高了雷达的点迹处理能力,在极限情况下能正确接收并处理雷达回波数据,满足了雷达改造需求。 展开更多
关键词 COM EXpreSS模块 回波 预处理 PCI EXpreSS总线 VXWORKS操作系统 数据处理
下载PDF
A Detection Method of Earthquake Precursory Anomalies Using the Four-Component Borehole Strainmeter 被引量:4
5
作者 Xiangyang Kong Kaizhi Su +1 位作者 Fujinawa Yukio Noda Yoichi 《Open Journal of Earthquake Research》 2018年第2期124-140,共17页
Strainmeters have been used to detect earthquake precursory anomalies in many countries. An innovated four-component strainmeter with four sensing units set at 45 degrees intervals, named SKZ strainmeter, was develope... Strainmeters have been used to detect earthquake precursory anomalies in many countries. An innovated four-component strainmeter with four sensing units set at 45 degrees intervals, named SKZ strainmeter, was developed and used in China. The design, with a few unique features, allows high-sensitivity monitoring of the regime of the crustal strain field, as well as the self-consistencies of the instrument. One of the most difficult problems in the earthquake precursory investigation is to efficiently detect anomalies from large amount of data. Pattern recognition of waveforms is widely used, but it is time-consuming and relies more or less investigator’s experience and decision. In this study, the consistency factors of the paired components were firstly shown to be utilized to detect anomalies possibly related with imminent earthquakes. Here, rather than using the consistency factors, the correlation coefficients of the two orthogonal strain data were used to detect. SKZ strainmeters have been installed at more than ten sites in China, exhibited high efficiency and reliability in precursory monitoring since. Anomalous variations from a few stations during two recent earthquakes in south China were analyzed. During normal stages, diurnal earth tides could be clearly observed with very little urban noises. Though the consistency factors may have near constant bias, their correlation coefficients remain near 1.0, greater than 0.99. During the imminent preparatory stage of earthquake occurrence, non-planar strain may appear and the correlation coefficients drop noticeably. The analysis showed that the correlation coefficient between the two orthogonal components is a useful parameter in post-processing of the strain data to detect precursory anomalies. The resultant resolving power is shown to be some one-order larger compared with previous methods. 展开更多
关键词 BOREHOLE Strainmeter EARTHQUAKE OBSERVATION Technology pre-Seismic ANOMALIES data FILTERING
下载PDF
Channel Estimation and Data Transmission in Massive MIMO TDD Systems 被引量:1
6
作者 Youjie Kang Tianyang Lu 《International Journal of Communications, Network and System Sciences》 2019年第10期170-187,共18页
This paper mainly elaborates the studies of channel estimation and downlink data transmission in Massive MIMO. As there are different types of interference in single-cell and multi-cell systems, this paper establishes... This paper mainly elaborates the studies of channel estimation and downlink data transmission in Massive MIMO. As there are different types of interference in single-cell and multi-cell systems, this paper establishes different models for them separately. In terms of uplink training, for getting channel state information, we introduce LS and MMSE channel estimation algorithms and make a comparison between them. At the same time, the problem of pilot contamination is solved by cell classification and pilot identification. Next, this paper defines mathematical models for downlink data transmission. We use pre-coding methods (including Zero-forcing and Maximal Ratio Combining schemes) and optimize power distribution to improve channel capacity and transmission rate. Furthermore, this paper provides numerical results to show the simulation performance in both single-cell and multi-cell systems and extends to prospects in the future. 展开更多
关键词 Channel Estimation PILOT CONTAMINATION DOWNLINK data Transmission pre-Coding Schemes Power Distribution
下载PDF
External calibration of GOCE data using regional terrestrial gravity data 被引量:4
7
作者 Wu Yunlong Li Hui +1 位作者 Zou Zhengbo Kang Kaixuan 《Geodesy and Geodynamics》 2012年第3期34-39,共6页
This paper reports on a study of the methodology of external calibration of GOCE data, using regional terrestrial-gravity data. Three regions around the world are selected in the numerical experiments. The result indi... This paper reports on a study of the methodology of external calibration of GOCE data, using regional terrestrial-gravity data. Three regions around the world are selected in the numerical experiments. The result indicates that this calibration method is feasible. The effect is best with an accuracy of scale factor at 10 -2 level, in Australia, where the area is smooth and the gravity data points are dense. The accuracy is one order of magnitude lower in both Canada, where the area is smooth but the data points are sparse, and Norway, where the area is rather tough and the data points are sparse. 展开更多
关键词 satellite gravity gradiometer GOCE external calibration terrestrial gravity data pre-process
下载PDF
数值预报中气象卫星资料同化前处理技术进展 被引量:1
8
作者 马刚 黄静 +5 位作者 巩欣亚 希爽 薛蕾 李娟 张鹏 龚建东 《应用气象学报》 CSCD 北大核心 2024年第2期142-155,共14页
在数值天气预报变分同化中,利用同化前处理将卫星资料完成有效信息优选、资料拼接和稀疏化、初级通道选择、下边界参数耦合等处理,实现卫星资料同化对数值天气预报业务的正贡献,是决定海量卫星资料同化效率、质量和效果的重要环节。针... 在数值天气预报变分同化中,利用同化前处理将卫星资料完成有效信息优选、资料拼接和稀疏化、初级通道选择、下边界参数耦合等处理,实现卫星资料同化对数值天气预报业务的正贡献,是决定海量卫星资料同化效率、质量和效果的重要环节。针对多种格式的卫星资料,中国气象局研发标准格式的高时效卫星资料拼接等技术,有效减小整轨卫星资料时间滞后对数值天气预报业务的负面影响。对于风云气象卫星资料,将云和降水检测、资料质量分析等处理置于同化前处理中,实现多光谱资料融合的同化预质量控制,保证了风云卫星微波温度探测资料和红外高光谱资料的同化正贡献。利用统一资料格式对预处理卫星资料进行再处理,拓展针对卫星成像和主动探测资料的处理,将卫星资料同化的部分质量控制功能置于卫星资料同化前处理中,是风云卫星资料同化前处理技术发展的重要趋势。 展开更多
关键词 数值天气预报 气象卫星资料 同化前处理
下载PDF
可撤销属性加密的区块链数据访问控制方法 被引量:4
9
作者 李健 戚湧 《计算机工程与设计》 北大核心 2024年第2期348-355,共8页
针对区块链数据共享中存在的粗粒度访问控制问题,提出一种基于属性撤销密文策略属性基加密的区块链数据访问控制方法。在现有方案基础上进行改造,引入预解密过程,结合属性撤销列表实现属性实时撤销;基于非对称群下的DBDH困难问题假设进... 针对区块链数据共享中存在的粗粒度访问控制问题,提出一种基于属性撤销密文策略属性基加密的区块链数据访问控制方法。在现有方案基础上进行改造,引入预解密过程,结合属性撤销列表实现属性实时撤销;基于非对称群下的DBDH困难问题假设进行安全性证明;基于超级账本Fabric进行系统设计,结合星际文件系统采用链上链下存储方式解决区块链容量不足和系统效率问题。实验结果表明,所提方案撤销属性时无需更新密钥密文重复上链,仅需要6次Pairing操作进行预解密和解密,且在大规模属性集下,预解密时间和解密时间平均保持在百毫秒左右的常量级上,实现区块链数据高效、细粒度的访问控制。 展开更多
关键词 区块链 数据共享 访问控制 属性基加密 预解密 属性撤销 星际文件系统
下载PDF
Measuring the Intraday Jump Tail Risk of Financial Asset Price with Noisy High Frequency Data
10
作者 Chao Yu Xujie Zhao Feng Zhang 《Open Journal of Statistics》 2017年第1期72-83,共12页
This paper proposes a simple two-step nonparametric procedure to estimate the intraday jump tail and measure the jump tail risk in asset price with noisy high frequency data. We first propose the pre-averaging thresho... This paper proposes a simple two-step nonparametric procedure to estimate the intraday jump tail and measure the jump tail risk in asset price with noisy high frequency data. We first propose the pre-averaging threshold approach to estimate the intraday jumps occurred, and then use the peaks-over-threshold (POT) method and generalized Pareto distribution (GPD) to model the intraday jump tail and further measure the jump tail risk. Finally, an empirical example further demonstrates the power of the proposed method to measure the jump tail risk under the effect of microstructure noise. 展开更多
关键词 High Frequency data Intraday JUMP Microstructure Noise JUMP TAIL Risk pre-Averaging
下载PDF
On Numerical methods for determination of Earth gravity field model using mass satellite gravity gradiometry data
11
作者 Zhu Guangbin Chang Xiaotao +2 位作者 Li Xinfa Zhang Xinhang Li Yuxing 《Geodesy and Geodynamics》 2012年第1期57-62,共6页
On the basis of Space-Wise Least Square method, three numerical methods including Cholesky de- composition, pre-conditioned conjugate gradient and Open Multi-Processing parallel algorithm are applied into the determin... On the basis of Space-Wise Least Square method, three numerical methods including Cholesky de- composition, pre-conditioned conjugate gradient and Open Multi-Processing parallel algorithm are applied into the determination of gravity field with satellite gravity gradiometry data. The results show that, Cholesky de- composition method has been unable to meet the requirements of computation efficiency when the computer hardware is limited. Pre-conditioned conjugate gradient method can improve the computation efficiency of huge matrix inversion, but it also brings a certain loss of precision. The application of Open Multi-Processing parallel algorithm could achieve a good compromise between accuracy and computation efficiency. 展开更多
关键词 satellite gravity gradiometry Cholesky decomposition pre-conditioned conjugate gradient open multi-processing parallel algorithm data processing
下载PDF
Emotion Deduction from Social Media Text Data Using Machine Learning Algorithm
12
作者 Thambusamy Velmurugan Baskaran Jayapradha 《Journal of Computer and Communications》 2023年第11期183-196,共14页
Emotion represents the feeling of an individual in a given situation. There are various ways to express the emotions of an individual. It can be categorized into verbal expressions, written expressions, facial express... Emotion represents the feeling of an individual in a given situation. There are various ways to express the emotions of an individual. It can be categorized into verbal expressions, written expressions, facial expressions and gestures. Among these various ways of expressing the emotion, the written method is a challenging task to extract the emotions, as the data is in the form of textual dat. Finding the different kinds of emotions is also a tedious task as it requires a lot of pre preparations of the textual data taken for the research. This research work is carried out to analyse and extract the emotions hidden in text data. The text data taken for the analysis is from the social media dataset. Using the raw text data directly from the social media will not serve the purpose. Therefore, the text data has to be pre-processed and then utilised for further processing. Pre-processing makes the text data more efficient and would infer valuable insights of the emotions hidden in it. The preprocessing steps also help to manage the text data for identifying the emotions conveyed in the text. This work proposes to deduct the emotions taken from the social media text data by applying the machine learning algorithm. Finally, the usefulness of the emotions is suggested for various stake holders, to find the attitude of individuals at that moment, the data is produced. . 展开更多
关键词 data pre-processing Machine Learning Algorithms Emotion Deduction Sentiment Analysis
下载PDF
数字孪生淠河建设实践与思考 被引量:1
13
作者 陈胜 李京兵 +3 位作者 王玉丽 张锦堂 刘昌军 刘业森 《中国水利》 2024年第13期50-56,共7页
数字孪生淠河根据淠河流域特征,以流域防洪需求为牵引,提出技术解决方案。在共享安徽省现有数据感知、网络传输、算力环境等基础上,建设内容包括流域数据底板、水利模型平台、水利知识平台等支撑平台,以及在上述支撑平台基础上建立防洪... 数字孪生淠河根据淠河流域特征,以流域防洪需求为牵引,提出技术解决方案。在共享安徽省现有数据感知、网络传输、算力环境等基础上,建设内容包括流域数据底板、水利模型平台、水利知识平台等支撑平台,以及在上述支撑平台基础上建立防洪预报、预警、预演、预案“四预”应用。防洪“四预”应用以精细化洪水预报为预警、调度提供基础,洪水预警为下游圩区等保护对象提供多阶段预警,情景化洪水预演以实景化方式分析洪水影响,智能化方案实现防洪预案的快速、自动化生成。通过总结数字孪生淠河建设内容,进一步探讨数字孪生流域建设中存在的难点问题,为相关数字孪生流域建设提供借鉴。 展开更多
关键词 淠河 数字孪生流域 圩区 流域数据底板 水利模型平台 水利知识图谱 “四预”应用
下载PDF
基于K-means SMOTE和IDBO-RF岩爆烈度等级预测模型 被引量:1
14
作者 温廷新 王泽锋 《中国安全生产科学技术》 CAS CSCD 北大核心 2024年第6期140-146,共7页
为解决岩爆数据集不均衡和模型参数寻优困难等问题,提出1种基于K-means SMOTE与改进蜣螂算法优化随机森林(random forest,RF)的预测模型。首先,分析岩爆发生机理构建指标体系;其次,使用K-means SMOTE算法对岩爆数据集进行均衡化处理,采... 为解决岩爆数据集不均衡和模型参数寻优困难等问题,提出1种基于K-means SMOTE与改进蜣螂算法优化随机森林(random forest,RF)的预测模型。首先,分析岩爆发生机理构建指标体系;其次,使用K-means SMOTE算法对岩爆数据集进行均衡化处理,采用Robust标准化消除量纲;最后,引入Tent混沌映射和非线性递减策略组合改进蜣螂优化(improved dung beetle optimizer,IDBO)算法,寻优RF超参数,建立岩爆烈度等级预测模型(IDBO-RF)并与其他模型对比验证其有效性。研究结果表明:数据均衡处理后,各模型准确率提高10.85%~16.02%;设计的IDBO-RF预测模型平均准确率约为94.37%,较RF、GWO-RF、DBO-RF模型分别提高约7.76百分点、1.69百分点、1.11百分点;IDBO-RF预测模型准确率最高约为96.43%,优于RF、GWO-RF、DBO-RF模型。研究结果可为解决岩爆预测问题提供一定参考。 展开更多
关键词 数据均衡 改进蜣螂优化(IDBO) 随机森林 岩爆烈度等级 预测模型
下载PDF
面向科研专病库的可视化前置数据采集系统建设探索与实践
15
作者 李哲明 朱珠 +2 位作者 黄坚 施珊珊 俞刚 《中国卫生信息管理杂志》 2024年第1期69-74,共6页
目的解决在基于临床数据中心构建专病库的过程中,由于数据字段频繁更新、手动补录工作量大导致的临床使用依从性不高的问题。方法提出了一种可视化前置数据采集的解决方案,采用MVC三层架构和WebService服务同步数据,支持可视化自定义病... 目的解决在基于临床数据中心构建专病库的过程中,由于数据字段频繁更新、手动补录工作量大导致的临床使用依从性不高的问题。方法提出了一种可视化前置数据采集的解决方案,采用MVC三层架构和WebService服务同步数据,支持可视化自定义病例报告表单,将采集动作前置到业务系统中即采即用,保障数据及时性,定期进行质量控制,保证数据来源的准确性。结果该方案在心内畸形和神母细胞瘤科研专病库上的应用显示,专病库数据质量和使用满意度得到显著提升,单份病例数据的整理时间从30分钟缩短到10分钟。结论该方案可减轻临床医生的工作负担,加快数据积累进度,提高科研专病库数据质量,激发临床医生对专病研究的科研热情。 展开更多
关键词 临床数据中心 前置数据采集 专病库 病例报告表单 数据质量
下载PDF
面向多源异质遥感影像地物分类的自监督预训练方法
16
作者 薛志祥 余旭初 +5 位作者 刘景正 杨国鹏 刘冰 余岸竹 周嘉男 金上鸿 《测绘学报》 EI CSCD 北大核心 2024年第3期512-525,共14页
近年来,深度学习改变了遥感图像处理的方法。由于标注高质量样本费时费力,标签样本数量不足的现实问题会严重影响深层神经网络模型的性能。为解决这一突出矛盾,本文提出了用于多源异质遥感影像地物分类的自监督预训练和微调分类方案,旨... 近年来,深度学习改变了遥感图像处理的方法。由于标注高质量样本费时费力,标签样本数量不足的现实问题会严重影响深层神经网络模型的性能。为解决这一突出矛盾,本文提出了用于多源异质遥感影像地物分类的自监督预训练和微调分类方案,旨在缓解模型对于标签样本的严重依赖。具体来讲,生成式自监督学习模型由非对称的编码器-解码器结构组成,其中深度编码器从多源遥感数据中学习高阶关键特征,任务特定的解码器用于重建原始遥感影像。为提升特性表示能力,交叉注意力机制模型用于融合异源特征中的信息,进而从多源异质遥感影像中学习更多的互补信息。在微调分类阶段,预训练好的编码器作为无监督特征提取器,基于Transformer结构的轻量级分类器将学习到的特征与光谱信息结合并用于地物分类。这种自监督预训练方案能够从多源异质遥感影像中学习到刻画原始数据的高级关键特征,并且此过程不需要任何人工标注信息,从而缓解了对标签样本的依赖。与现有的分类范式相比,本文提出的自监督预训练和微调方案在多源遥感影像地物分类中能够取得更优的分类结果。 展开更多
关键词 遥感 多源异质数据 预训练 自监督学习 土地覆盖分类
下载PDF
大庆SN油田东部过渡带油水边界综合确定
17
作者 梁宇 杨会东 +3 位作者 付宪弟 蔡东梅 王彦辉 孙衍民 《新疆石油地质》 CAS CSCD 北大核心 2024年第2期213-220,共8页
为确定大庆SN油田东部过渡带油水界面,综合钻井、测井、地震资料,结合岩心含油产状分析和老井油水层二次解释,基于双相介质理论的叠后属性油气检测以及基于叠前地震波形指示反演的流体识别等技术,探讨构造油气藏外扩区油水边界的综合确... 为确定大庆SN油田东部过渡带油水界面,综合钻井、测井、地震资料,结合岩心含油产状分析和老井油水层二次解释,基于双相介质理论的叠后属性油气检测以及基于叠前地震波形指示反演的流体识别等技术,探讨构造油气藏外扩区油水边界的综合确定方法。研究区油水边界具有以下特征:岩心含油产状为油斑以上;测井解释外推为油层或油水同层;叠后属性低高频能量比大于0.85;叠前反演预测含水饱和度小于75%。因此,以“井点定深度、地震定边界、动态来验证”为原则,从“点—线—面—空”,经过综合分析,确定最终油水边界位置。研究成果有效指导了研究区外扩评价部署,对同类型构造油田油水边界研究具有参考意义。 展开更多
关键词 构造油气藏 油水边界 井震资料 叠后属性 叠前反演 综合分析
下载PDF
医疗大模型技术及应用发展研究
18
作者 陈晓红 刘浏 +3 位作者 袁依格 王俊普 李大元 邱建华 《中国工程科学》 CSCD 北大核心 2024年第6期77-88,共12页
医疗大模型基于深度神经网络架构进行复杂医疗数据的高效处理与模式识别,为智慧医疗提供新型的决策支持;需要系统分析医疗大模型技术及应用情况,以精准把握医疗大模型的发展方向、精准应对面临的发展挑战,进而基于医疗大模型提升医疗文... 医疗大模型基于深度神经网络架构进行复杂医疗数据的高效处理与模式识别,为智慧医疗提供新型的决策支持;需要系统分析医疗大模型技术及应用情况,以精准把握医疗大模型的发展方向、精准应对面临的发展挑战,进而基于医疗大模型提升医疗文本、医学图像、药械研发、医学教育等方面的能力。本文梳理了医疗大模型的技术范式与应用场景,剖析了由基础层、模型层、应用层、公共模块构成的医疗大模型技术体系,覆盖评价指标体系构建、数据集范围与题型、模型对齐方法、模型评测平台的医疗大模型评测体系,辨识出医疗大模型应用存在的数据安全、技术风险、落地挑战、伦理道德等方面的难点。为此建议,发挥政府引导优势、保障数据安全,加快基础理论研究、突破技术风险,强化应用场景牵引、缓解落地挑战,建立健全监管机制、规范伦理道德,完善公共服务体系、营造创新生态,以加快医疗大模型创新应用,推动我国智慧医疗的高端化、智能化、绿色化发展。 展开更多
关键词 医疗大模型 多模态数据 预训练微调 提示工程 技术体系 评测体系
下载PDF
基于院前120急救大数据的流行病学特征及其空间分析——以大连市为例
19
作者 李鹏 邱琬钧 +4 位作者 马楠 郭建科 刘永睿 张帅 伍美强 《辽宁师范大学学报(自然科学版)》 CAS 2024年第2期235-242,共8页
城市公共卫生应急管理与居民生命安全息息相关,随着大连市人口老龄化速度加快,疾病谱系也随之复杂化,居民对急救的需求与日俱增.采用2020年9月至2021年8月大连市120院前急救患者的历史大数据,筛选并分析院前急救患者的流行病学特征建立... 城市公共卫生应急管理与居民生命安全息息相关,随着大连市人口老龄化速度加快,疾病谱系也随之复杂化,居民对急救的需求与日俱增.采用2020年9月至2021年8月大连市120院前急救患者的历史大数据,筛选并分析院前急救患者的流行病学特征建立数据库,利用最近邻距离指数、核密度、缓冲区分析等方法研究大连市120院前急救患者的空间格局分布的类型和特征.通过筛选符合标准的40 551病例后,对院前急救患者进行流行病学分析,得到大连市120院前急救患者的流行病学特征:男女比例为1.13∶1,疾病高峰年龄段是老年组(病发年龄≥60岁),排除不明确诊断,病种占比前3的疾病是损伤和中毒(24.727%)、心血管系统(10.219%)、呼吸系统(9.928%),且这些疾病都主要发生在老年组.院前120急救患者在空间上多呈现集聚分布,主要集中于苍山路周围、友谊广场附近、东联路与华北路相交一带、香炉礁街、联合路两侧、中山广场附近、人民路、二七广场周围、马栏北街、西山小区附近、台山街周围.以期对指导医院急救资源合理配置以及制定有效的院前急救防控措施具有一定的参考意义. 展开更多
关键词 流行病学 院前120急救 急救患者大数据 GIS空间分析
下载PDF
基于半监督学习的域适应实体解析算法
20
作者 戴超凡 丁华华 《计算机科学》 CSCD 北大核心 2024年第9期214-222,共9页
实体解析旨在查找两个数据实体是否引用同一实体,是许多自然语言处理任务中的一项基本任务。现有的基于深度学习的实体解析解决方案通常需要大量的标注数据,即使利用预训练的语言模型进行训练,仍然需要数千个标签才能达到令人满意的准... 实体解析旨在查找两个数据实体是否引用同一实体,是许多自然语言处理任务中的一项基本任务。现有的基于深度学习的实体解析解决方案通常需要大量的标注数据,即使利用预训练的语言模型进行训练,仍然需要数千个标签才能达到令人满意的准确性。现实场景中,这些标注数据并不容易获得。针对上述问题,提出了一个基于半监督学习的域适应实体解析模型。首先,在源域上训练一个分类器,然后利用域适应减小源域和目标域的分布差异,同时用数据增强后的目标域软伪标签加入源域迭代训练,从而实现从源域到目标域的知识迁移。在13个来自相同或不同领域的数据集上对所提模型进行了对比实验和消融实验,实验结果表明,与无监督基线模型相比,所提模型在多个数据集上的F1值平均提升了2.84%,9.16%和7.1%;与有监督基线模型相比,所提模型只需要20%~40%的标签就可以达到与有监督学习相当的性能。消融实验进一步证明了所提模型的有效性,其总体上可以获得更好的实体解析结果(相关代码已开源1))。 展开更多
关键词 实体解析 域适应 伪标签 预训练语言模型 数据增强
下载PDF
上一页 1 2 45 下一页 到第
使用帮助 返回顶部