应用BP神经网络预测嵌岩桩竖向承载力的研究

利用BP神经网络,建立了嵌岩桩竖向承载力预测的BP神经网络模型,采用嵌岩桩竖向承载力的实验数据对网络进行了训练,并以此训练好的BP神经网络对嵌岩桩承载力进行了预测。预测结果表明,该模型具有很高的预测精度,为预测嵌岩桩承载力提供了一种新的方法。     

环境振动的灰色预测模型研究

利用 GIM( 1 )的非时序直接建模法来预测研究建筑施工的环境振动 ,并将 GIM( 1 )模型与 GM( 1 ,1 )模型进行比较分析。结果表明 GIM( 1 )模型的拟合精度优良 ,对原始资料中白化信息的利用更加丰富 ,拓宽了 GIM( 1 )模型在环境科学领域中的应用范围     

铁路专用线声环境影响评价

针对新建铁路专用线沿线的环境特征，提出了铁路专用线声环境影响评价的程序、内容和方法。在目前缺乏新建铁路噪声预测模式的情况下，提出了应对铁路边界噪声、不同距离处的铁路噪声、敏感点噪声、鸣笛噪声进行预测，并以某铁路专用线声环境影响评价为例，对上述四种噪声进行了预测，通过预测得到铁路两侧的声环境状况，并以此为依据提出了铁路噪声的防治对策，以保证铁路专用线沿线声环境质量。     

城市空气质量数值预报系统对PM2.5的数值模拟研究

发展了南京大学城市空气质量数值预报模式系统(NJU-CAQPS),在模式系统中引入了气溶胶模块.运用该系统对南京地区冬夏两季PM2.5浓度的时间变化规律和空间分布特征进行了数值模拟研究,通过与实际观测资料对比,检验发展后的模式系统对于细颗粒物的模拟性能.结果表明,南京市城近郊内冬夏两季PM2.5浓度具有明显的时空变化特征,一般在半夜和清晨会出现较高浓度,午后至傍晚浓度较低.冬季浓度高于夏季,冬夏两季算例的浓度日均值之比为1.51.空间分布受到排放源位置、地面流场等因素影响.二次气溶胶在PM2.5中占相当的份额,冬夏两季算例中二次气溶胶在PM2.5中所占比例分别为12%和15%,夏季二次气溶胶对PM2.5浓度贡献较冬季大.与实际观测资料的对比验证表明,经过发展的该模式系统对于城市PM2.5等颗粒物的模拟性能良好.     

常压塔塔顶腐蚀关键参量相关性分析与预测

预测常压塔顶回流罐切水的关键离子浓度可以为常压塔顶系统工艺防腐提供技术指导。收集了某炼厂2014-2016年常压装置的生产实时工艺参数、原料及产品采样数据、水质分析化验参数,采用线性插值方法对不同时间尺度的数据进行补全。通过线性相关性分析获得了影响常压塔顶回流罐切水总铁离子浓度、pH的主要因素,并基于深度学习、支持向量机回归、粒子群优化方法建立了腐蚀关键参量预测模型。结果表明,pH值和总铁离子浓度的相关因素大部分重叠,与原料性质及产品馏出温度较强相关;建立的回归模型预测精度高,在训练集和预测集上,总铁离子浓度预测值与测量值最大偏差分别为4. 4%和9. 8%,pH值预测值与测量值最大偏差分别为0. 9%和1. 1%。     

Adaptations to the dynamics of rural water supply from natural sources: A village example in semi-arid Nigeria

Domestic and agro-economic activities in the semi-arid region of Nigeria rely exclusively on rainfall, streamflow and groundwater in deep and alluvial (fadama) aquifers. Such water supply systems are subject to considerable seasonal and inter-annual variability. However, a combination of the various sources may mitigate the effects of water scarcity. This study describes the dynamics of the water sources for a village in northeastern Nigeria as an example that demonstrates the linkages between rainfall, streamflow and groundwater. Such linkages are important for developing strategies to mitigate the effects of climatic variation. Long-term records of rainfall were not available at the site so that the short-term data was interpreted in the context of the long-term climatic experience of the region. The stratigraphic profile of the aquifer was developed from resistivity methods, to supplement information obtained from well level hydrographs. Field observations and analysis of the data reveal that the beginning of the rainy season is a precarious period because none of the water sources are reliable. Rainfall infiltration recharges the shallow aquifers while lateral water flux from the floodwater-saturated fadama also contributes to water level rises in the deep wells. A study of the water level changes in response to deepening of the wells suggests that the wells are fed by an assortment of water-bearing lenses, separated by layers of low transmissivity. One major finding of this study is that there exists a close association between the wells in this riparian community and the recharge processes within the fadama. Upstream diversion, abstraction or impoundment could therefore undermine the viability of aquifer exploitation by the communities in the riparian zone. Analysis of the interactions among the various sources of water leads to the identification of several potential adaptation strategies for confronting the problem of water scarcity.     

Yellow River Valley flood and drought disaster:spatial-temporal distribution predictionand early-warning

ＹｅｌｌｏｗＲｉｖｅｒＶａｌｌｅｙｆｌｏｏｄａｎｄｄｒｏｕｇｈｔｄｉｓａｓｔｅｒ：ｓｐａｔｉａｌ－ｔｅｍｐｏｒａｌｄｉｓｔｒｉｂｕｔｉｏｎｐｒｅｄｉｃｔｉｏｎａｎｄｅａｒｌｙ－ｗａｒｎｉｎｇＧａｏＬｉｎ，ＳｈａＷａｎｙｉｎｇ，ＬｉｕＨｕａｉｑｕａｎ，...     

对美国FHWA公路噪声预测模式的深入讨论

美国ＦＨＷＡ公路噪声预测模式在我国被广为应用，甚至有各种修正模式，但因对原模式的理论依据仍缺少了解，以至在实际应用过程中或对原模式进行修正时诸如参数的选择及计算、模式适用范围等出现了不当或错误现象。本文从理论上对ＦＨＷＡ公路噪声预测模式进行了深入分析、进而讨论了各参数的物理意义、模式的适用范围，并提供了二种特殊情况下公路噪声的预测模式。     

Prenatal diagnosis of spinal muscular atrophy in Russia

Ninety-two families with spinal muscular atrophy (SMA) applied for genetic counselling and further prenatal diagnosis. To minimize expenses, only one tightly linked informative marker was determined in the course of preliminary examination, and non-radioactive allele detection was preferably used. Four prenatal diagnoses of SMA type I, four of SMA type II, and one of SMA type III were made. This trial programme shows the considerable requirements, importance, and potential effectiveness of prenatal prediction of SMA in Russia.     

深圳地铁罗湖车站深基坑位移监测结果分析

依据基坑监测的实际数据，通过分析桩锚支护结构的水平位移挠曲线，论述了锚索的锚固效果及相应的改进方法，总结桩背土体深层位移的特点和发展趋势，同时利用灰色模型对支护结构进行了位移预测、比较，说明了其准确性。