基于桥连接的词典学习方法的语义解析

现阶段的语义解析方法大部分都基于组合语义,这类方法的核心就是词典。词典是词汇的集合,词汇定义了自然语言句子中词语到知识库本体中谓词的映射。语义解析一直面临着词典中词汇覆盖度不够的问题。针对此问题,该文在现有工作的基础上,提出了基于桥连接的词典学习方法,该方法能够在训练中自动引入新的词汇并加以学习,为了进一步提高新学习到的词汇的准确度,该文设计了新的词语—二元谓词的特征模板,并使用基于投票机制的核心词典获取方法。该文在两个公开数据集(WebQuestions和Free917)上进行了对比实验,实验结果表明,该文方法能够学习到新的词汇,提高词汇的覆盖度,进而提升语义解析系统的性能,特别是召回率。     

一种新型微处理器功能验证

文中介绍一种新型微处理器的验证方法，并对相关的一些验证策略进行深入的讨论。     

TTI hydraulics and erosion control laboratory research field performance of erosion-control blankets

Storm water management issues facing the Texas Department of Transportation in the late 1980s led to the development of a coordinated research program with the Texas Transportation Institute. Researchers developed methodologies for evaluating the field performance of various erosion control technologies of the most widely used products within the Department's construction and maintenance operations. From these methodologies, the Hydraulics and Erosion Control Laboratory was designed and constructed. Currently, participants include private industry (manufacturers of erosion control products), transportation researchers (TTI), and the public sector (TxDOT).The results reported in this paper reflect 2 years of erosion-control blanket research. The study objectives were to determine the effectiveness of erosion-control blankets on the growth of warm-season perennial grasses and their ability to prevent sediment loss in a sloped condition.The laboratory simulates the highway environment with the sloped plots (6 m in width) located on an earthen embankment that is 300 m in length and 6.75 m in height (94 ft. by 22 ft., vertically). A randomized experimental design was replicated on two soil types (sand and clay) for each slope condition (3:1 or 2:1) with a control.In general, the results indicate better combined results relating to sediment retention and vegetation establishment performance for erosion-control blankets on sandy soils (noncohesive) regardless of slope condition (3:1 or 2:1) or material type. A minimum of 50% more sediment was retained on the sandy treatment plots and a 45% more vegetation coverage was achieved compared with the control plots. When analyzed by material type related to performance, excelsior, synthetic blends, and straw/coconut blends performed the best. For clay soils (cohesive), regardless of slope condition (3:1 or 2:1), the combined results indicate a minimum of 75% more sediment was retained and a minimum of 5% more vegetation establishment was achieved compared with the control plots. When analyzed by material type related to performance, excelsior, straw, and straw/coconut blends performed the best.     

基于人工势场方法的WSN传感覆盖率提高算法

感覆盖率是WSN系统中重要的性能指标之一。Sink节点的移动性能提高系统的传感覆盖率,人工势场能描述普通节点与Sink节点之间的关系。提出了一种利用人工势场方法有效的Sink节点移动控制策略,通过Sink节点的移动能够提高系统的传感覆盖率。普通节点搜集的数据能通过多跳的方式传送给Sink节点。在最初的随机部署之后,由人工势场产生的力迫使移动Sink节点移动到整个环境中的平衡点以获得更高的传感覆盖率。在整个算法执行过程中无需事先了解环境的情况,此方法能在实时的环境中采用。给出了具体的算法步骤,最后利用一个仿真实例对控制效果进行了说明。     

栅栏覆盖最大化生存周期调度算法

栅栏覆盖是近年来无线传感器网络的研究热点之一,如何延长生存周期是无线传感器网络研究的一个重要问题。针对无线传感器网络的栅栏覆盖应用,设计了两种最大化网络生存周期的调度算法:集中式的Greedy调度算法和分布式的DBCS调度算法。仿真实验表明:两种算法显著地延长了网络的生存周期;在较稀疏的网络中,DBCS算法与Greedy算法性能接近,分布式的DBCS算法适合应用于大规模传感器网络。     

2000—2014年辽宁省植被动态演变特征分析

基于MODIS － NDVI 数据,辅以线性回归法与分段线性回归法,并借助ArcGIS 软件,对辽宁 省2000—2014 年植被覆盖的动态演变过程进行分析。结果表明: ( 1) 时间上,辽宁省植被NDVI 在年 际尺度上呈现出明显的增大趋势,2005 年出现突变,多年平均NDVI 值为0. 496; 春季、夏季、秋季 以及植被生长季NDVI 突变年份分别为2006 年、2005 年、2009 年和2004 年,秋季波动变化的突变点 明显滞后; 植被生长最旺盛的季节为夏季,且集中于8 月。( 2) 空间上,辽宁省植被覆盖具有明显的 地域性差异,呈现出东部高、中西部低的分布特征; 辽宁省植被覆盖优良区与辽东山地的界限基本吻 合,植被覆盖贫乏区主要集中在朝阳市和阜新市的东北部地区。( 3) 辽宁省植被覆盖程度呈山地阴坡 高于阳坡的形态,并且植被覆盖程度最好的坡向为北偏西方向。( 4) 2000—2014 年辽宁省植被覆盖度 整体以维持现状和轻微改善为主,保持不变的区域集中于中东部地区,辽阳市与沈阳市一带有轻微退 化现象,辽西北地区改善情况较为明显。     

陕北无定河流域土壤侵蚀与植被覆盖和降雨关系研究

基于2000—2014年陕北无定河流域日降水、DEM、土壤类型、MODIS NDVI等数据,利用修正通用土壤流失方程(RUSLE)估算了流域土壤侵蚀情况,定量分析了植被覆盖和降雨在土壤侵蚀中的作用,结果表明:(1)无定河流域土壤侵蚀以微度侵蚀为主,平均占流域面积的88.35%,各土壤侵蚀等级面积比例随等级的升高而降低。2000—2014年微度侵蚀面积比例为下降趋势,轻度及以上等级面积比例均为上升趋势。(2)不考虑降雨因子影响时,无定河流域仍以微度侵蚀为主,且为增加趋势,而其他土壤侵蚀等级均为下降趋势。不考虑植被覆盖因子时,流域土壤侵蚀与模型计算结果的变化趋势基本一致。(3)植被覆盖对土壤侵蚀面积比例和变化速率的贡献率分别为13.67%和24.55%,而降雨作用达到86.33%和75.44%,表明降雨是流域土壤侵蚀的主要动力和控制因子,降雨变化主导着流域土壤侵蚀变化过程。     

Consequences of different cutting regimes on regrowth and nutrient stoichiometry of Sparganium erectum L. and Potamogeton natans L.

Aquatic vegetation forms an essential component in freshwater ecosystems but due to changed environmental and anthropogenic conditions often needs management to reduce nuisance for human land‐use. In this paper, the authors looked at the regrowth of two macrophyte species (Potamogeton natans and Sparganium erectum) in two lowland rivers under different cutting treatments. After an initial cross‐sectional transect was manually removed from bank to bank at the beginning of the growth season, a monthly repetitive removal of biomass in plots on that transect was done during the rest of the growth season (testing frequency of mowing). Additional new transects were also cut in subsequent months (testing timing of mowing). Finally, biomass was repetitively removed in plots in those additional transects too (testing frequency of mowing × timing of mowing). The biomass at the end of the growth season was analysed for C, N, P, and Si. It was demonstrated that timing and frequency of vegetation cutting has an important effect on the capacity and rate of species' recovery and therefore on the efficiency of the applied management. Nutrient stoichiometry of the regrown biomass was directly affected by cutting. Caused by differences in the applied timing and frequency of the cutting, C/N and N/P ratios and BSi concentrations were highly variable. Yet, overall, there was a clear tendency towards a higher C/N ratio and BSi concentration and lower N/P ratio in biomass that recovered after cutting. This human impact on the quantity and quality of autochthonous organic matter may have knock‐on effects on the decomposers food web and mineralization process.     

无位置信息的无线传感器网络漏洞发现算法*

无线传感器网络广泛应用的前提是能够检监测目标区域特定事件的发生。而节点能量受限、难补给的特点导致监测性能难以保证,造成监测盲区出现。针对无线传感器网络节点的地理位置信息未知的情况进行了研究,提出覆盖漏洞发现算法CHDA (Coverage Holes Discovery Algorithm),在以节点为极点建立的极坐标中计算出相邻节点的相对位置信息和该节点被其邻居节点覆盖的边缘弧的信息,并且根据节点的单纯覆盖弧序列的定义计算出节点被其邻居节点覆盖的单纯覆盖弧序列,从而得到网络中的覆盖漏洞。覆盖漏洞的发现克服的地理位置信息未知的限制,为漏洞的修补提供了必要的前提条件进而保证传感器网络的覆盖率。     

加筋生态护坡技术的应用与发展

在越来越注重工程生态性的今天,生态护坡在道路、堤防、城市水利等方面得到了广泛应用。然而很多情况下,生态护坡在强度上无法满足工程设计的要求。加筋生态护坡是土工织物与植草相结合形成的一种护坡形式,在保证工程生态性的同时大大提高了生态护坡的强度,近年来逐渐得到重视与应用。综合介绍了加筋生态护坡的发展历程,将其概括为传统生态护坡、普通加筋生态护坡和高性能加筋生态护坡3个阶段,在此基础上对加筋生态护坡使用的不同结构型式和材质的土工合成材料进行了分类总结,针对国内外加筋生态护坡技术的研究现状和主要结论,结合实例分析了加筋生态护坡的工程应用,指出在保护边坡、减少水土流失及改善生态环境方面的作用。最后,展望了未来需要研究的方向。