黑龙江省单一及混作农作物种植结构提取研究

以2012年MODIS数据产品250 m精度的MOD13Q1数据为数据源,采用Savitsky-Golay滤波法提取并平滑NDVI时间序列曲线,提取11种物候关键值,利用AMP,Start,Length三种物候关键值,采用面向对象分类方法的隶属度函数法提取出黑龙江省农用地范围,然后选取AMP,Start,Length,Left,Base,Right,Small七种物候关键值,采用面向对象分类方法的标准最邻近法,提取出黑龙江省单一及混作农作物的种植结构共十种。     

自适应动态演化粒子群算法在Web主题信息搜索中的应用

针对传统的基于单一价值评价的网络爬虫搜索策略存在的不足,提出了一种基于自适应动态演化粒子群(adaptive dynamical evolutional particle swarm optimization,ADEPSO)的启发式网络爬虫搜索算法。本算法综合立即价值和未来价值两种链接评价方法,并依据链接价值所反映的Web实际搜索情况动态调整两种价值的关系,使网络爬虫能更准确地预测页面的重要性。实验表明,该算法具有较高的搜索效率。     

基于多分类器组合面向对象遥感影像信息提取方法

提出最近距离法和基于知识规则的模糊分类法相结合的混合分类法,针对IKONOS遥感影像,分别用最近距离法、基于知识规则的模糊分类法以及混合分类法对影像进行信息提取。结果表明：混合分类法的信息提取精度最高,总体精度提高到95.60%,Kappa系数提高到0.944,其为面向对象的高分辨率影像信息提供理想方法。     

Prediction of hollows in abandoned underground workings at shallow depth

Composite statistical analysis of the lithological composition of the rock mass above underground workings in coal seams and experimental work (laboratory mechanical tests and boring for the determination of hollows) in Donetsk city (Ukraine) have been used to develop a method for predicting the presence of hollows. The empirical criteria K₁–K₂>0 and K₃2.5 which relate to the physical characteristics of the overlying strata were found to predict the existence of such hollows in abandoned underground workings at shallow depth. The values of K₁–K₂ and K₃ depend on the thicknesses of the different rock layers and the uniaxial compressive strength of the immediate roof over underground openings. The different layers e.g. sandstones, argillites, aleurolites and alluvium in the rock mass are shown to influence the existence of hollows in abandoned workings. Large thicknesses of sandstones in the rock mass or relatively high uniaxial compressive strength rock in the immediate roof contribute to the existence of hollows in abandoned workings. On the other hand, sandstones of small thickness, weak alluvium, argillites and aleurolites only give additional weight on immediate roof of the opening and allow collapse of the rock mass. The method of prediction for hollows was developed from the back-analysis of data from 41 boreholes which were drilled from the ground surface above underground openings.     

A Mssbauer effect study of Fe3+ bearing

Three synthetic Fe3+ bearing γ-Fe2SiO4 were analyzed usingelectron probe method, and the Mssbauer spectra of the samples at 298 K, 150 K, and 95 K were measured. Each spectrum at three temperatures is composed of two doublets. These two doublets are assigned to Fe2+ in the octahedral sites and Fe3+ in the tetrahedral sites, respectively. Site occupancies were determined. The results show that Fe3+ and a small amount of Si4+ are in the tetrahedral and octahedral sites, respectively. The average bond lengths of the octahedral and tetrahedral sites were calculated according to the equations primarily given by Hill et al., O’Neill and Navrotsky and modified by the authors. Furthermore, the octahedral and tetrahedral bond lengths were used to calculate cell parameters and oxygen parameters. In addition, Fe3+ line broadening in the M?ssbauer spectra of Fe3+ bearing γ-Fe2SiO4 were interpreted by using the next nearest neighbor effects.     

基于Quaternary编码的球面三角格网邻近搜索算法(英文)

The characteristic of Quaternary codes is analyzed. The rule of distinguishing triangle direction is given out. An algorithm of neighbor finding by decomposing the Quaternary code from back to front is presented in this paper. The contrastive analysis of time complexity between this algorithm and Bartholdi＇s algorithm is approached. The result illustrates that the average consumed time of this algorithm is about 23.66% of Bartholdi＇s algorithm.     

隧道断面截取曲面拟合优化的点云简化方法

为了提高变形监测中地铁隧道断面点截取的效率,文章提出了基于kd-tree和法向量估计的局部点云简化方法,对BaySAC算法的三维激光点云二次参数曲面拟合方法进行改进:利用kd-tree建立点云数据的空间拓扑关系,计算出每个数据点的k邻域;然后使用平面拟合方法获取法矢量;最后根据点云数据法矢量变化程度,采用法矢量自适应得到压缩后的点云数据。实验证明该方法既能较大程度地简化点云,简化结果比较均匀,又具有不破坏细小特征的特点,进一步改进了BaySAC算法的二次参数曲面拟合方法。     

浑善达克沙地南缘榆树种群的大小结构和邻体格局

 浑善达克沙地作为京津地区沙尘暴的重要沙源地引起了生态学界的广泛关注。榆树是浑善达克沙地的优势种,榆树疏林在浑善达克沙地成缀块状分布,对于该区域生态系统的整体性和多样性维持以及防风固沙等方面具有重要意义。研究浑善达克沙地南缘榆树种群的大小结构和邻体格局的结果显示:榆树种群的平均胸径为25.40 cm,最小胸径为6.37 cm,缺乏幼树,种群更新不良。榆树邻体格局研究显示:榆树种群呈丛分布,取样个体到最近邻体距离的平均值是7.13 m,在4 m以下的占55.0%,10 m以下的占72.6%。两个个体互为最近邻体的榆树占了71.4%,具有成对现象,成对现象对榆树有明显影响。最后对榆树疏林存在的问题及保护措施等进行了探讨。     

自适应的IDW插值方法及其在气温场中的应用

反距离权重(Inverse Distance Weighting,IDW)插值通常采用距离搜索策略选择插值参考点,当采样点集分布不均匀时,距离搜索策略使得参考点聚集一侧影响插值精度。自然邻近关系具有良好的自适应分布特性,可有效地解决参考点分布不均匀问题。结合自然邻近关系,提出自适应的反距离权重(Adaptive-IDW,AIDW)插值方法。首先对采样数据构建初始Delaunay三角网,然后采用逐点插入法,将待插值点插入初始Delaunay三角网中,局部调整得到新的Delaunay三角网,以待插值点的一阶邻近点作为IDW插值的参考点,使参考点自适应均匀地分布在待插值点周围,再进行IDW插值计算。利用AIDW插值方法对Franke函数、全国气温观测数据进行插值实验,结果表明此方法具有较高的精度,且减少了“牛眼”现象。     

基于WIFI的位置指纹参考节点密度与定位精度关系研究

为探究基于WIFI的位置指纹室内定位中定位精度与离线数据库参考节点密度关系,本文采用待测节点与参考节点位置关系原理理论推导、正方形、三角形和六边形三种不同参考节点的排布方式的计算机拟合并以正方形排布方式为例进行实际验证的方法,对参考节点密度与定位精度关系开展研究。回归分析结果表明,参考节点密度与定位精度在不同排布方式下均呈现显著正相关线性关系。