网络分解法的误差分析

网络分解法是求解电大尺寸复杂结构电磁问题的重要方法之一,关于网络分解法误差分析的研究有着重要的理论意义和实用价值.本文给出了网络分解法的误差估计,并对于脉冲基函数转换成三角基函数的情况给出了三角基函数个数的最优公式,数值结果表明了分析的正确性.     

城市中心区游憩功能的开发——以武汉市为例

从我国游憩空间建设实践中 ,概括出对城市游憩功能认识过程的四个阶段 .进而 ,从城市中心商业区、历史街区和城市滨水区三个方面讨论了城市中心区的游憩功能开发问题 .最后 ,以武汉市为例进行了分析     

Assessment of Capillary Rise from Shallow Groundwater by the Simulation Model SIMWASER Using Either Estimated Pedotransfer Functions or Measured Hydraulic Parameters

The performance of the water balance and crop growth model SIMWASER to estimate the water supply of a lawn by capillary rise from shallow ground water was investigated by using lysimeter measurements with a sandy and a clayey soil. Moreover the robustness of the model was evaluated by running it either with measured hydraulic soil parameters or with estimated pedotransfer functions (PTFs) derived from texture and bulk density. Simulations were performed for the years 1996–1998 with input data from the lysimeter station Berlin-Dahlem (Germany) for lysimeters containing undisturbed sandy or clayey soil monoliths, with groundwater at 135 cm depth. Simulated evapotranspiration and percolation/capillary rise were in good agreement with the measured data for all variants, while simulated soil water storage was acceptable only for the variants using hydraulic soil data based on laboratory measurements or using PTFs derived from known soil class and bulk density. PTFs based on mean total pore volume of the respective soil classes yielded soil water storages which were evidently too high; therefore they should be used with care and must be avoided at all in simulating the soil water balance of arid sites with shallow groundwater.     

CARS光谱测温及仪器狭缝函数对测量的影响

设计了一种利用光参量振荡器(OPO)来代替染料激光器的宽带相干反斯托克斯喇曼散射(CARS)光谱测温实验装置。为了提高CARS测量温度的精度,测量和分析了检测系统中狭缝函数对测量结果的影响,确定了狭缝函数的具体形式。将测量结果与热电偶读数比较,发现在温度范围300K～1200K内,温度测量的不确定度小于2%。     

用改进的逐次逼近解法计算二维井间电磁场

提出了一种计算积分方程的新方法——改进的逐次逼近解法 (MSAM) ,并用该方法计算了轴对称二维井间电磁场。与传统的逐次逼近解法 (SAM)相比 ,该方法收敛性强 ,应用范围广 ,可适用于高电导率对比地层。由于不必进行直接的大型矩阵求逆运算 ,因此与积分方程的直接解法 (IE)相比 ,该方法计算速度更快 ,所需内存更少。采用矩阵求逆方法计算了成层介质中的二维 Green函数 ,并对含 Green函数的积分进行了简化 ,从而加快了计算速度。数值计算结果显示 ,在地层电导率对比度达到 2个数量级时 MSAM仍收敛 ,且计算结果与直接求解积分方程的结果一致 ,因此 MSAM是一种有效的计算轴对称二维井间电磁场的方法     

Phase separation studies in RIM polyurethanes catalyst and hard segment crystallinity effects

Polyurethanes were prepared from pure 4,4′-diphenylmethane diisocyanate (MDI), 1,4-butane diol (BDO) or 1,2-ethane diol (EDO) and α,ω-hydroxyl poly(propylene oxide) (PPO) by reaction injection moulding (RIM). Hard segment (MDI + BDO or EDO) level was 45–50 wt%. The PPO had about 20% ethylene oxide copolymerized in at the chain ends to provide 80% primary OH end groups. M_n was varied from 2000 to 4000. Dibutyl tin dilaurate catalyst and mould temperature were varied. Dynamic mechanical, wide-angle X-ray, differential scanning calorimeter, molecular weight and tensile elongation measurements were made on the RIM polyurethanes. At low reaction rates (low catalyst or temperature) highly crystalline, well phase separated but low molecular weight polymers were produced. At high catalyst or temperature levels more poorly phase separated but high molecular weight, tough polymers resulted. Higher M_n PPO gave better phase separation and EDO gave higher melting temperatures. Preventing hard segment crystallinity by substituting asymmetric MDI or glycols resulted in phase compatibility.     

变异函数在进间砂体预测中的应用

应用变异函数探讨了砂体的定量预测问题。通过对北二东密井网试验区不同类型砂体，不同井网密和岩厚度的预测，检验了此方法的精度（为70％左右），在原方法的基础上，通过对预报精度的分析并结合精度地质研究进一步完善了该方法，提出二次建模法并应用于北二东西块聚驱井的砂岩厚度预测中，预测精度可提高5％以上，从而说明应用变异函数研究砂体的定量预测问题具有可借鉴性，可以为今后新井调整厚度的预测及砂体描述提供依据。     

超高分子量聚乙烯衬板在煤仓中的应用

介绍了井底煤仓溜煤斜面采用超高分子量聚乙烯衬板安装及应用情况 ,对井底煤仓棚煤原因及解决办法进行了探讨     

TRANSFER FUNCTION MODEL ORDER AND PARAMETER ESTIMATION

Abstract. A distributed lag model with a rational transfer function is considered. We have demonstrated that the C -table can be used to determine the orders of the rational transfer function and the Padé table can be used to estimate the parameters of an identifiable transfer function. In case the model is not uniquely identified, the C -table reduces possible competing models to only a few for further examination. Padé approximant gives consistent initial values for possibly more efficient iterative procedures to estimate the coefficients of the identified model. Some asymptotic results on the estimation of the C -table and the Padé table are given. A few examples are presented to demonstrate the effectiveness of the method.