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。     

人工神经网络在地质学判别分析中的应用

将现代人工神经网络方法应用于地质学的判别分析，通过实例计算表明，无论是对于两组判别分析还是多组判别分析，人工神经网络方法都是切实可行的。     

Ultimate strength of ship hulls under torsion

For a ship hull with large deck openings such as container vessels and some large bulk carriers, the analysis of warping stresses and hatch opening deformations is an essential part of ship structural analyses. It is thus of importance to better understand the ultimate torsional strength characteristics of ships with large hatch openings. The primary aim of the present study is to investigate the ultimate strength characteristics of ship hulls with large hatch openings under torsion. Axial (warping) as well as shear stresses are normally developed for thin-walled beams with open cross sections subjected to torsion. A procedure for calculating these stresses is briefly described. As an illustrative example, the distribution and magnitude of warping and shear stresses for a typical container vessel hull cross section under unit torsion is calculated by the procedure. By theoretical and numerical analyses, it is shown that the influence of torsion induced warping stresses on the ultimate hull girder bending strength is small for ductile hull materials while torsion induced shear stresses will of course reduce the ship hull ultimate bending moment.     

Standard facies models to incorporate all heterogeneity levels in a reservoir model

The construction of reservoir models is frustrated by the fact that core and well cover only a fraction of the reservoir volume and it is therefore difficult to determine features like facies shape, -size, and -distribution, inter- and intra-facies boundaries and lateral trends from them. These features are, however, critical to fluid flow and they should necessarily be incorporated in the reservoir model and we therefore propose to systematically describe geometry and distribution of facies. To this end we make use of “standard facies models” that a priori contain all elements and boundaries of facies for a number of typical depositional environments.     

Prediction of Load Carrying Capacity of Corroded Reinforced Concrete Beam

A novel method for prediction of the load carrying capacity of a corroded reinforced concrete beam (CRCB) is presented in the paper. Nine reinforced concrete beams, which had been working in an aggressive environment for more than 10 years, were tested in the laboratory. Comprehensive tests, including flexural test, strength test for corroded concrete and rusty rebar, and pullout test for bond strength between concrete and rebar, were condueted. The flexural test results of CRCBs reveal that the distribution of surface cracks on the beams shows a fractal behavior. The relationship between the fractal dimensions and mechanical properties of CRCBs is then studied. A prediction model based on artificial neural network (ANN) is established by the use of the fractal dimension as the corrosion index, together with the basic intbrmation of the beam. The validity of the prediction model is demonstrated through the experimental data, and satisfactory resuits are achieved.     

沉降监测多项式回归分析与神经网络预测

使用多项式和切比雪夫(Tchebyshev)多项式分别对沉降监测数据进行回归分析以预测未来沉降值,其中切比雪夫多项式的外推效果较好;应用前向BP神经网络对两种不同的单因子输入模式进行非线性函数逼近,并进行了不同采样步长的比较,实例表明将时间点作为网络的输入对沉降进行预测效果较好。     

Response-based method for determining the extreme behaviour of floating offshore platforms

The accurate prediction of extreme excursion and mooring force of floating offshore structures due to multi-variete environmental conditions which requires the joint probability analysis of environmental conditions for the worst case situation is still impractical as the processing of large amount of met-ocean data is required. On the other hand, the simplified multiple design criteria (e.g. the N-year wave with associated winds and currents) recommended by API known as traditional method does lead neither to the N-year platform response nor to the N-year mooring force. Therefore, in order to reduce the level of conservatism as well as uncertainties involved in the traditional method the response-based method can be used as a reliable alternative approach. In this paper this method is described. In order to perform the calculations faster using large databases of sea states, Artificial Neural Networks (ANN) is designed and employed. In the paper the response-based method is applied to a 200,000 tdw FPSO and the results are discussed.     

人工神经网络方法反演东中国海悬移质浓度

海中悬移质是决定海洋光学性质、海洋水质,河口海岸带演变动力过程的重要环境参数。本文利用模拟遥感反射比数据集建立人工神经网络反演悬移质浓度,并利用东中国海现场同步数据对该算法进行验证。