基于支持向量机方法的森林火险预测研究

以利用气象数据预测森林火险等级为目的,使用半正定规划建模选定支持向量机的最优核函数,依据500余条林火数据建立了支持向量回归机模型,并使用回归误差特征曲线图对比分析各个回归模型的学习效果。分析得到该自定义核函数的均方误差为1.76,支持向量数为190,约占训练数据集的1/2。结果表明,与传统的线性回归方法及基于高斯核的支持向量机相比,该预测模型具有较高的准确率并且有效的避免了过学习的现象。     

Penetration behavior of bath induced by top-blown gas jets for steelmaking converter

The cavity induced by top-blow gas jets impinging on the bath surface is one of most important parameters, which has overwhelming influence on the rate of slag formation and metallurgical performance. Water model experiments are conducted through a 1/10 scaled-down top-blown convertor. Effects of lance height and gas flow rate on penetration behavior are studied, and also the influence mechanism of liquid surface tension on penetration depth is addressed based on the experiments and theory analysis. An energy utilization index (EUI) is defined to reflect how much the impinging kinetic energy of gas jets are transferred into bath for formation of cavity, and an improved model is proposed to predict penetration depth of jets. The results show that penetration depth increases with lowering lance height and increasing gas flow rate, and impact diameter increases with increasing lance height, but is little affected by gas flow rate. Influence of liquid surface tension on penetration depth is enhanced with increasing surface tension and penetration depth. The EUI increases with increasing lance height, and a function relationship of it with lance height is obtained based on the experiments.     

植物叶面积系数法改进研究

植物叶面积计算是植物野外调查的重要关键内容,由于传统植物叶面积计算公式s=kab中,k值随物种的不同而存在很大差异,往往造成公式应用的困难。为了方便k值的正确选择,收集了具有代表性的9种形态的植物叶片,求出不同形态叶片的k值,为植物叶面积野外调查提供方便。k值计算方法为,利用计算机软件Image Pro Plus 6.0计算了植物叶片的面积、面积与外接矩形的面积比、最大直径、最小直径、宽度与长度共6个参数。借鉴传统植物叶片的分类方法,根据最大直径与最小直径的比值,将叶片分为披针形、长卵形、卵圆形、矩形或圆形4种类型,求出这4种类型叶片的k值分别为1/3,1/2,2/3和5/6。经过试验证明,公式在改进前,叶片面积k值的平均误差接近20%,改进后叶片面积的k值的平均误差仅为4%,叶片面积的估测误差明显降低。按叶形将叶片分类确定k值后,k值的选择方便快速,外业人员只需简单培训就可明显提升野外作业的效率和精度。     

The predictive control of an incremental model based on tracking error

An algorithm of incremental model predictive control is proposed to solve a class of industrial production process real-time tracking problems. The basic idea is to establish a predictive model by adopting speed response on the basis of tracking error, and realize the optimization of the second time performance by using the soft constraint of the increment of control quantity and the online rolling optimization method. The theoretical analysis and Matlab simulation results show that the control algorithm can obtain good tracking performance, smooth dynamic response of the controlled object and strong robustness of the closed-loop system.     

Serum 5'nucleotidase activity in rats: a method for automated analysis and criteria for interpretation

A manual kit for determining serum 5'nucleotidase (5'NT, EC 3.1.3.5) activity was adapted for use with rat samples on a large discrete clinical chemistry analyzer. The precision of the method was good (within-run C.V. = 2.14%; between-run C.V. = 5.5%). A comparison of the new automated method with a manual and semi-automated method gave regression statistics of y = 1.18X -3.66 (Sy. x = 4.54), and y = 0.733X + 1.97 (Sy. x = 1.69), respectively. Temperature conversion factors provided by the kit manufacturer for human samples were determined to be inaccurate for converting results from rat samples. Analysis of components contributing to normal variation in rat serum 5'NT activity showed age and sex to be major factors. Increased serum 5'NT activity was observed in female rats when compared to male rats beginning at about 5 to 6 weeks of age. An analysis of variance of serum 5'NT, alkaline phosphatase, and GGT activities observed over a 9-week period in normal rats suggests several advantages for 5'NT as a predictor of biliary lesions in rats.     

孔隙结构图像分析中不同试验因素对分析结果的影响

研究了不同试验因素 (包括图像分辨率、土壤切片定向性、分析区域大小等 )对孔隙结构图像分析结果的影响以及试验误差 ,并简要介绍了用于分析土壤孔隙结构的土壤切片及数字图像的制备技术。结果表明 :不同试验因素对分析结果均会产生一定的影响 ,试验误差约在 1 0 %。这表明同一研究中应保持试验条件的相对一致以保证分析结果的可靠性和可比性     

不同样本尺度和分区方案的粮食产量空间化及误差修正

粮食产量数据空间化有助于粮食产量数据与其他自然、人文数据进行综合分析,但空间化过程中必然会产生误差。该文按照3种分区方案(全国不分区、全国分为7个区以及按省分区),选择3种尺度上(县级、地市级和公里网格)的总产及平均产量数据(即4种样本:县级粮食总产、县级平均粮食产量、地市级粮食总产、地市级平均粮食产量)分别为因变量,以对应的3种农田类型(水田、水浇地、旱地)面积数据为自变量,利用多元线性回归分析方法,得到15种空间化模型。采用两阶段误差分析方法,选取2个模型误差评价因子和5个空间化结果误差评价因子,对模型和空间化结果进行误差分析。结果表明:1)空间化过程中,模型精度与空间化结果的精度存在不一致性;2)对于采用同一样本的模型(常数项为0)而言,空间化结果精度随着分区方案的细化先提高再降低,而对于采用同一样本的模型(常数项非0)而言,空间化结果精度随着分区方案的细化而降低;3)在全国不分区和分为7个区2种情况下,空间化结果精度随着分析样本尺度的细化(从地市级到县级再到公里网格)先提高后降低。根据上述分析结果,最终以县级粮食总产为样本、常数项为0、全国分7个区建模的方案实现全国粮食产量数据空间化,并通过修正,得到2005年中国粮食产量公里网格分布图。该研究弥补了粮食产量空间化误差分析的不足,探寻了不同样本尺度和分区方案与空间化误差的关系,提高了空间化精度,同时对其他类型的社会经济统计数据空间化研究具有一定的参考价值。     

滑柱摆臂式悬架车辆之转向梯形断开点的确定

根据机构学原理,将目前农用运输车采用的滑柱摆臂式悬架等效成铰接四杆机构;通过实例计算和断开点主要影响因素分析,提出了正确选择转向梯形断开点位置的新方法。     

土地整理综合效益评价指标与方法

土地整理综合效益评价包括经济效益、社会效益和生态效益三方面，根据多次专家咨询和问卷调查结果，逐次求取影响土地整理综合效益评价的指标，用层次分析法(AHP)确定各评价指标权重，量化各评价指标，求出综合效益评价值，并以江苏省溧阳市新昌镇土地整理综合效益评价为例。新昌镇土地整理综合效益评价结果表明：该镇土地整理综合效益评价效果较好，但社会效益和部分生态效益指标增加不明显，在土地整理实践中应重视促进土地整理的经济效益、社会效益和生态效益协调发展。     

林冠截留降雨研究综述

林冠截留降水是植被对降水到达地面的第一次阻截,也是对降雨的第一次再分配,减少了林地的有效降雨量,是水文循环过程的重要环节。从林冠截留量及影响因素、林冠截留模型和与之相关的林冠覆盖度、叶面积指数的确定方法等方面论述了目前国内外的研究现状,分析当前植被截留研究的发展趋势。