基于Mindlin理论Winkler地基参数的Kalman滤波识别

提出了Winkler地基参数的Kalman滤波识别方法。引入Mindlin理论后,推导了Winkler地基上板的控制微分方程。利用Fourier变换,推求了Winkler地基上简支板的Fourier闭式解。推导了Winkler地基参数的Kalman滤波方程,研究了Winkler地基参数的Kalman滤波识别的具体计算步骤。研究表明,运用Kalman滤波理论进行Winkler地基参数的识别,能有效地估计Winkler地基参数;Winkler地基参数的滤波收敛速度和精度受地基参数初始值选取和位移实测资料的影响,且Kalman滤波理论也可用于其他地基模型地基参数的识别。     

钱二块铀矿床采铀注液结垢趋势的理论分析与预测

用离子系数矩阵法对钱二块铀矿床地浸采铀试验注液中存在的独立离子反应进行了确定,由溶度积规则及反应平衡原理,借助Matlab编程,估算出了钱二块铀矿床地浸采铀试验注液可能会生成的沉淀物种类及数量,为防治结垢提供了一定的理论依据.     

基于AGA的SVM需水预测模型研究

需水预测是一个由城市人口、工业水平、社会经济水平共同作用的多因素、多层次的复杂非线性系统.其结果将直接影响受区域水资源承载力约束的产业结构、布局形态等决策.作为一种集中参数预报方法,支持向量机方法具有对未来样本的较好的泛化性能,对于这类资料缺乏、系统结构尚欠清晰的问题可以取得较好的模拟和预测结果.基于此,本文将支持向量机方法引入需水预测领域,建立了需水预测支持向量机模型.同时,本文将加速遗传算法和支持向量机方法耦合起来,构造了支持向量机模型参数的自适应优化算法.模型在珠海市的应用实例表明:与简单遗传算法比较,AGA的模型参数寻优效率更高;与BP神经网络模型相比,SVM模型较好地解决了小样本、经验性等问题,并取得了较高的预测精度.     

东海区近海强风预报模式

本文探讨东海区近海强风预报模式，首先介绍建立预报模式的思路和步骤，及使用说明，为了验证强风预报模式和判别条件，用１９８６、１９８７上日本传真天气图进行试报，准确和基本准确度为９１．１％，并给出了１９８９、１９９５年随船试验预报结果，准确和基本准确率为９０％以上。     

嵌入式Linux技术在海洋自动观测系统中的应用研究

随着Linux技术逐步走向成熟，嵌入式Linux技术在海洋自动观测系统中的应用将全面提高海洋观测系统整体性能，降低系统开发成本，进一步推进我国海洋自动观测系统产权自主化的建设。文章就如何研制适用于海洋自动观测系统的嵌入式Linux系统进行了论述。     

浙北地区旱涝预报问题探讨—一种汛期降水量长期预报方法

本文分析了当地前期气象要素——温度的演变特点与汛期降水量之间的关系,建立了两者之间的回归方程,并由此作出降水量的长期预报,从预报误差和试报结果看,效果是好的。     

烟威渔场盐度预报及分析

本文以我国北黄海的烟威渔场为研究对象,采用优选因子场预报模式,对盐度的时空变化进行预报试验。文中对所采用的统计预报方法,作了简要介绍。对影响本海区盐度变化的显著因子及预报结果作了初步分析。预报试验表明:预报的总体平均绝对误差为0.27‰,预报相对误差在18％左右,预报误差小于0.5‰的站数占总站数的85％,预报趋势与实测资料基本一致。     

Siltation Prediction for Navigation Channels and Harbour Basins on Muddy Beach

- This paper, after briefly reviewing the experimental research on sediment transport on muddy beach since the 1950s, improves and perfects the method for forecasting siltation in navigation channels and harbour basins which was first put forward in China by the authors. In consideration of silty sediment and sand, some factors in forecasting methods have been changed and modified. Consequently, the modified methods can be used either to compute siltation in navigation channels and harbour basins on muddy beach or to compute siltation and scouring in navigation channels and harbour basins on both silty beach and sandy beach. The verification of field data from eleven large, medium and small natural harbours shows a good agreement between the forecasting by the modified method and the natural conditions. Finally, the paper deals with the rational utilization of water area after the construction of the West Dyke in Lianyungang, the maintenance of water depth of the navigation channel at the entrance, siltation distribution, siltation in the navigation channel and harbour basin for ships of 100 thousand tonnnage. Results once again prove that the prospect of constructing Lianyungang Harbour into a deepwater harbour is bright.     

Fatigue Behavior of Plain Concrete Under Biaxial Compression: Experiments and Theoretical Model

The effects of different lateral confinement stress on the fatigue behavior of and cumulative damage to plain concrete are investigated experimentally. Eighty 100 mm x 100 mm x 100 mm specimens of ordinary strength concrete are tested under constant- or variable-amplitude fatigue loading and lateral confinement pressure in two orthogonal directions. A fatigue equation is developed by modifying the classical Aas-Jakobsen S-N equation for taking into account the effect of the confined stress on fatigue strength of plain concrete. The results of variable-amplitude fatigue tests indicate that the linear damage theory proposed by Palmgren and Miner is unreasonable in the biaxial stress state. A nonlinear cumulative damage model that could model the effects of the magnitude and sequence of variable-amplitude fatigue loading and lateral confinement pressure is proposed on the basis of the evolution laws of the residual strains in the longitudinal direction during fatigue tests. The residual fatigue life predic     

一种基于BP算法学习的小波神经网络

为发展 Szu的基于信号表示的小波神经网络 ,提出一种多输入多输出的小波网络模型 ,网络隐层采用框架小波函数、输出层采用 Sigmoid激励函数 ,并选用“熵误差函数”以加速网络的学习速度。奇偶判别和混沌时间序列预测例子的实验结果表明了它具有良好的函数逼近能力和推广能力 ,收敛速度和均方误差均优于相同结构的多层感知器模型。