组合优化的能源消费量预测模型

针对非等间距灰色系统预测中存在误差较大的问题，结合序列本身的特点，利用世界能源消费的历史数据，采用3种灰色预测模型与神经网络进行组合优化，建立了灰色神经网络的能源消费量组合预测模型。实证分析结果表明，提高了模型的拟合和预测精度，拓宽了应用范围。该模型可对能源的消费趋势进行预测，为科学分析能源结构提供依据。     

BEL型双维振打式电除尘器的研发

文章介绍了一种顶打和侧打相结合的BEL型双维振打式电除尘器，该产品将电除尘器两种主要的振打清灰技术结合，适用于老除尘器的提效改造及新除尘器的制造等。     

低周荷载下型钢混凝土柱抗剪性能智能分析

为了研究型钢混凝土短柱的抗震性能，通过21根SRC柱在低周反复荷载作用下的剪切黏结破坏试验结果，在利用灰色关联理论进行模型输入变量选择的基础上，建立了其基于改进神经网络模型的受剪承载力分析模型，进而研究了混凝土抗压强度、腹板宽度、轴压比及箍筋间距等因素对SRC柱抗剪能力的影响规律，并对SRC柱的受剪承载力进行了预测。结果表明，所提方法能够反映受剪承载力与影响因素间的非线性变化规律，可供SRC柱在低周往复荷载作用下的受剪承载力分析参考。     

三峡茅坪溪沥青混凝土心墙堆石坝位移反演分析与安全预测

本文研究了基于综合应用人工神经网络和演化算法的位移反演分析方法,并将此方法应用于三峡茅坪溪沥青混凝土心墙堆石坝的变形反演分析中。以茅坪溪一期工程原型观测成果为依据,反演出能够正确反映坝体变形特性的邓肯-张EB模型参数,从而预测了施工期末和蓄水期末该坝的变形特性,并基于总应力法研究了心墙水力劈裂破坏发生的可能性,为茅坪溪堆石坝运行期的安全生产提供依据。研究结果表明,茅坪溪心墙堆石坝不会发生水力劈裂破坏,该坝蓄水后虽然水压上升,但心墙仍是安全的。     

多因素耦合条件下硫化矿自燃神经网络动态预测模型研究

硫化矿石自燃是多种因素、多场耦合综合作用的结果,是一典型的非线性问题。笔者应用人工神经网络技术,以Matlab软件为平台,通过现场调查和理论分析,建立了矿石含硫量、通风强度、环境温度3因素与硫化矿石自燃之间的预测模型;通过数据样本学习与部分现场监测数据相结合进行模拟,研究表明预测数据与实测结果基本吻合,误差控制在10%以内,取得了较好的效果。该研究为预防硫化矿石自燃提供一个新的思路和方法,具有一定的理论意义和应用价值。     

CALCULATION OF CUMULATIVE PERCENTAGE AT SELECTED PARTICLE SIZES FROM BOTTOM WITHDRAWAL TUBE ANALYSIS

ABSTRACT: A new computation method is developed for the bottom withdrawal tube particle size analysis. A transformation on time scale is introduced so that the cumulative percentage of a selected particle size can be determined from a single continuous Oden cum.     

人工神经网络在水环境质量评价中的应用

为了将人工神经网络应用于水环境质量评价，应用了人工神经网络Ｂ—Ｐ算法，构造了水环境质量评价模型，该模型应用于实例评价结果表明，人工神经网络用于环境质量评价具有客观性，通用性和实用性。     

Genetic amniocentesis: 505 cases performed before the sixteenth week of gestation

Between May 1987 and November 1988, 505 early amniocentesis within the 15th week of gestation were performed at the First Department of Obstetrics and Gynaecology,‘L. Mangiagalli’ of the University of Milan and at the Department of Obstetrics and Gynaecology of ‘Gaslini’ hospital in Genoa. A total number often abnormal fetal karyotypes were diagnosed. In addition, one case of pseudomosaicism (not confirmed on fetal blood) and one case of osteogenesis imperfecta type II (observed at ultrasound examination) were also detected. Eleven pregnancies were therefore terminated because of an abnormal fetus. Out of 494 pregnancies (excluding terminated pregnancies) there were 16 fetal losses within the 28th week; ten of these occurred in the 2 weeks following the procedure. There were 475 live-births, of which 447 were term deliveries and the other 28 deliveries occurred before the 37th week of gestation.     

The fetal behavioural states: An ultrasonic study

In order to accurately detect the fetal behavioural state, we simultaneously measured fetal heart rate and multiple fetal activities in 27 healthy pregnant women at 38 to 40 weeks of gestation. We ultrasonically identified gross body movements, breathing movements and micturition. Analysis of fetal heart rate allowed us to distinguish two different patterns of fetal behaviour: active and quiet phases. The frequency distribution of the analysed fetal events was significantly different in these two phases. These data suggest that a complete biophysical profile of the fetus is effective in differentiating behavioural states and may improve the predictive accuracy of fetal heart rate analysis alone.