采用STRAIGHT模型和深度信念网络的语音转换方法

提出一种将STRAIGHT模型和深度信念网络DBN相结合实现语音转换的方式。首先,通过STRAIGHT模型提取出源说话人和目标说话人的语音频谱参数,用提取的频谱参数分别训练两个DBN得到语音高阶空间的个性特征信息;然后,用人工神经网络ANN将两个具有高阶特征的空间连接并进行特征转换;最后,用基于目标说话人数据训练出的DBN来对转换后的特征信息进行逆处理得到语音频谱参数,并用STRAIGHT模型合成具有目标说话人个性化特征的语音。实验结果表明,采用此种方式获得的语音转换效果要比传统的采用GMM实现语音转换更好,转换后的语音音质和相似度与目标语音更接近。     

Predicting temporal centrality in Opportunistic Mobile Social Networks based on social behavior of people

Predicting the centrality of nodes is a significant problem for different applications in Opportunistic Mobile Social Networks (OMSNs). However, when calculating such metrics, current studies focused on analyzing static networks that do not change over time or using aggregated contact information over a period of time. Furthermore, the centrality measured in the past is not verified whether it is useful as a predictor for the future. In this paper, in order to capture the dynamic behavior of people, we focus on predicting nodes’ future centrality (importance) from the temporal perspective using real mobility traces in OMSNs. Three important centrality metrics, namely betweenness, closeness, and degree centrality, are considered. Through real trace-driven simulations, we find that nodes’ future centrality is highly predictable due to natural social behavior of people. Then, based on the observations in the simulation, we design several reasonable prediction methods to predict nodes’ future temporal centrality. Finally, extensive real trace-driven simulations are conducted to evaluate the performance of our proposed methods. The results show that the Recent Weighted Average Method performs best in the MIT Reality trace, and the recent Uniform Average Method performs best in the Infocom 06 trace. Furthermore, we also evaluate the impact of parameters m and w on the performance of the proposed methods and find proper values of different parameters for each proposed method at the same time.     

基于Tri-Training的事件关系分类方法研究

事件关系分类是一项研究事件之间存在何种逻辑关系的自然语言处理技术。针对事件关系分类任务中训练语料不足的问题,提出了基于Tri-Training的事件关系分类方法。该方法首先根据已标注的语料训练三个不同的分类器,以多数投票的方式从未标注集中抽取置信度较高的样本对训练集进行扩充,然后利用新的训练集重新训练分类器,反复迭代,不断完善分类模型,最终达到提升事件关系分类性能的目的。实验结果表明,以F1值为评价标准,基于Tri-Training的事件关系分类方法在四大类事件关系上的分类性能为64.36%。     

层次聚类分析方法在水土保持生态修复分区中的应用

本文主要从防治水土流失、促进植被恢复和生产经济发展的角度,以水土保持生态修复理论为指导,在平凉市市域范围内多因素复杂条件下应用层次聚类分析方法,研究水土保持生态修复分区并得到了较好的结果,为合理配置水土保持生态修复措施和投资提供了科学依据。     

On the Controllability of Nonlocal Second‐Order Impulsive Neutral Stochastic Integro‐Differential Equations with Infinite Delay

In this paper, we investigate the controllability for a class of nonlocal second‐order impulsive neutral stochastic integro‐differential equations with infinite delay in Hilbert spaces. More precisely, a set of sufficient conditions for the controllability results of nonlocal second‐order impulsive neutral stochastic integro‐differential equations with infinite delay are derived by means of the Banach fixed point theorem combined with theories of a strongly continuous cosine family of bounded linear operators. As an application, an example is provided to illustrate the obtained theory.     

水利工程竣工验收检测评价体系构建

通过调查分析全国有代表性的水利工程质量检测和竣工验收检测的技术资料,以及对工程建 设行业已有检测技术标准的对比分析和研究,提出了水利工程竣工验收检测分部和检测单元的划分、 检测项目和抽样数量、检测技术和评价方法,以及检测评价体系的构成,为制定水利工程质量验收检 测评价技术标准提供技术支持。     

石羊河流域水资源承载能力研究

基于石羊河流域经济社会发展和水资源利用状况,在分析水资源承载能力内涵的基础上,建 立了水资源承载能力计算方法,分析了石羊河流域的水资源承载状态,指出石羊河流域在当前和今后 一段时期仍将处于超载状态,需要继续加大水资源保护力度,促进实现经济社会可持续发展。     

WATER TUNNEL EXPERIMENTAL INVESTIGATION ON THE DRAG REDUCTION CHARACTERISTICS OF THE TRAVELING WAVY WALL

Drag reduction experiment of the traveling wavy wall at high Reynolds number is conducted.A suit of traveling wavy wall device is developed.The drag forces of the traveling wavy wall with various wave speeds(c)are measured under different water speeds(U)in the K15 cavitation water tunnel and are compared with that of the flat plate.The results show that the mean drag force of the traveling wavy wall have decreased and then increased with oscillation frequency increasing at the same flow speed.Under different flow speeds,when traveling wave wall reached to the minimum of drag force,the corresponding the ratio of the wall motion phase speed c to flow speed U,c /U is slightly different.Within the parameters of the experiment,whenc /U reaches a certain value,the drag force of the traveling wavy wall can be less than that of the flat plate.The drag reduction can be up to 42%.Furthermore,as the value ofc /U increases,the traveling wavy wall can restrain the separation and improve the quality of flow field.     

一种地震属性图像配色阈值选取算法

针对传统的地震属性图像可视化过程中存在的缺点,论文提出一种基于直方图技术的地震属性图像配色阈值选取算法,该算法利用直方图统计的结果,首先进行阈值预选,然后在区间内进行阈值调整,最后根据所确定的阈值对地震属性图像进行重新配色,很好地解决了传统地震属性图像视觉上的杂乱以及区域性差的缺点。经过对大庆油田多个区块地震属性图像的分析,该算法可有效解决地震图像可视化中的不足,提高了地震解释工作的效率。