基坑支护结构水平变形预测的遗传神经网络方法

采用遗传算法和误差反向传播算法相结合的混合算法来训练前馈人工神经网络,先用遗传学习算法进行全局训练,再用BP算法进行精确训练。就遗传算法过程中的选择、变异进行了探索,提出了用BP网络训练产生变异的遗传算法。作为实例,将该方法应用于预测基坑支护结构水平变形中。结果表明,该方法有收敛速度较快、预测精度高等优点。     

四川银矿床的成因类型及找矿方向

四川已发现银矿床35个，主要与铅、锌、铜共生或者为其伴生组份。笔者将四川银矿床初步划分为5个类型和9个亚类型。成矿时代以晚白垩世和晚三叠世为主。四川银矿的找矿方向主要是三江地区的德格—乡城深大断裂带和扬子地台西缘的攀西裂谷带。     

内蒙古胜利煤田共生锗矿的成因地球化学初探

用统计和作图的分析方法,对内蒙古胜利煤田中共生锗矿床的锗品位与挥发分、灰分、硫分的关系进行了研究。结果表明煤中锗品位与挥发分呈正相关、与原煤灰分负相关、与洗煤灰分正相关关系;全区煤属中硫煤,锗品位与原煤和洗煤硫分均呈正相关关系,煤的灰分指数较低(3 579);锗分布不均匀,富锗煤矿明显属于断裂坳陷边缘沉积,锗品位可以在煤层的顶部、中部、底部同时或单个部位呈现高值,大多数煤层中部出现锗高品位值,这有别于以往所报道的锗品位只在煤层顶、底部相对富集的研究结论。认为胜利煤田锗主要与有机质结合;锗的有利聚集条件是水动力较弱、地下水位较低、强还原的停滞沼泽环境;泥炭聚集阶段有机质的吸附作用是锗的主要富集机制。     

浅水方程数值计算方法的研究

求解以水位为变量的连续方程,并根据Navier-Stokes方程压力修正算法的基本思想,建立了浅水方程的水位修正算法,放宽了对离散时间步长的限制.通过对离散方程系数矩阵的重新构造,建立了高分辨率有限元格式,该格式既具有较高的离散精度又避免了数值解的伪振荡.对动量方程的阻力项做负坡线性化处理,提高了露滩计算的稳定性.数值模拟结果与解析解吻合良好,表明所建立的数值计算方法是正确的和可靠的.     

鄂尔多斯盆地苏里格庙与靖边天然气单体碳同位素特征及其成因

苏里格庙上古生界砂岩气藏天然气重烃含量高,干燥系数低,为湿气;靖边奥陶系风化壳气藏天然气重烃含量低,干燥系数高,属干气。苏里格庙天然气的正构烷烃碳同位素具有δ13 C₁<δ13 C₃ <δ13 C₂ <δ13 C4的规律,靖边天然气具有δ13 C₁<δ13 C₂ <δ13 C₃ 的规律。二者高碳数烷烃的同位素差别较大,尤其是苏里格庙的乙烷碳同位素比靖边平均重 3.6‰。根据天然气单体碳同位素及其含量变化,苏里格庙天然气属煤成气,靖边天然气属煤成气与油型气的混合气。苏里格庙和靖边天然气中均含少量CO₂ 气,其碳同位素表明苏里格庙除苏 5井为无机CO₂ 气外,均为有机CO₂ 气;靖边林 1、林 5井为无机CO₂ 气,陕参 1井为有机CO₂ 气。结合烷烃气和CO₂ 气,苏里格庙天然气可以分为含有机CO₂ 煤成气和含无机CO₂ 煤成气。碳同位素比值与Ro 关系分析认为,苏里格庙的天然气成熟度处于成熟到高成熟阶段,其相应气源岩的平均Ro 为 1.2 9%～ 1.39%,这一结果与该区二叠系煤岩的成熟度符合较好。     

The inversion of anelastic coefficient,source parameters and site respond using genetic algorithm

It gradually becomes a common work using large seismic wave data to obtain source parameters, such as seismic moment, break radius, stress drop, with completingof digital seismic network in China (Hough, et al, 1999; Bindi, et al, 2001). These parameters are useful on earthquake prediction and seismic hazard analysis.Although the computation methods of source parameters are simple in principle and the many research works have been done, it is not easy to obtain the parameters accurately. There are two factors affecting the stability of computation results. The first one is the effect of spread path and site respond on signal. According to the research results, there are different geometrical spreading coefficients on different epicenter distance. The better method is to introduce trilinear geometrical spreading model (Atkinson, Mereu, 1992; Atkinson, Boore, 1995; WONG, et al, 2002). In addition, traditional site respond is estimated by comparing with rock station, such as linear inversion method (Andrews, 1982), but the comparative estimation will introduce some errors when selecting different stations. Some recent research results show that site respond is not flat for rock station (Moya, et al, 2000; ZHANG,. et al, 2001; JIN, et al, 2000; Dutta, et al, 2001). The second factor is to obtain low-frequency level and corner frequency fromdisplacement spectrum. Because the source spectrum model is nonlinear function,these values are obtained by eye. The subjectivity is strong. The small change of corner frequency will affect significantly the result of stress drop.     

Coupled inverse modelling of groundwater flow and mass transport and the worth of concentration data

This paper presents the extension of the self-calibrating method to the coupled inverse modelling of groundwater flow and mass transport. The method generates equally likely solutions to the inverse problem that display the variability as observed in the field and are not affected by a linearisation of the state equations. Conditioning to the state variables is measured by an objective function including, among others, the mismatch between the simulated and measured concentrations. Conditioning is achieved by minimising the objective function by gradient-based methods. The gradient contains the partial derivatives of the objective function with respect to: log conductivities, log storativities, prescribed heads at boundaries, retardation coefficients and mass sources. The derivatives of the objective function with respect to log conductivity are the most cumbersome and need the most CPU-time to be evaluated. For this reason, to compute this derivative only advective transport is considered. The gradient is calculated by the adjoint-state method. The method is demonstrated in a controlled, synthetic study, in which the worth of concentration data is analysed. It is shown that concentration data are essential to improve transport predictions and also help to improve aquifer characterisation and flow predictions, especially in the upstream part of the aquifer, even in the case that a considerable amount of other experimental data like conductivities and heads are available. Besides, conditioning to concentration data reduces the ensemble variances of estimated transmissivity, hydraulic head and concentration.     

遗传算法与单纯形法组合的影像纹理分类方法

提出遗传算法(简称GA)与单纯形法组合的影像纹理分类方法(简称GASPX)。单纯形法是一种局部搜索方法，它通过反射，扩张，收缩操作，求得新的单纯形点，组成新单纯形，新单纯形比前一个单纯形更接近局部最优解。这种寻优方法收敛速度快，它与GA组合起来可以改善单独使用GA收敛速度慢的缺陷。由于在组合算法中是多个单纯形的局部区域的并行搜索，避免GA优化过程中过早收敛于局部最优解的现象出现。通过5种不同类别航空影像纹理识别的试验，并与GA的结果作对比，结果表明GASPX法优于GA法。     

指派交叉--一种新的遗传交叉方法

在用遗传算法生成影像纹理分类的“Tuned”模板中，引用了一个新的遗传交叉方法——指派交叉，与其他的交叉方法相比，指派交叉可在相同的搜索空间中提供更好的优化组合。试验证明，指派交叉是值得推广的一种遗传交叉方法。     

模拟退火算法在动态建立基准地价模型中的应用

阐述了传统基准地价评估模型建立流程将地价样本检验与模型建立分离的不足，引入了模拟退火算法思想，对目标函数、初始状态、新状态产生、新状态接受、初温、温度更新函数以及抽样准则、算法终止准则等关键问题进行设计，将地价样本检验、优化组合与模型建立有机结合，并设计了基于模拟退火算法的动态建立基准地价模型的流程，最后应用实例验证了方法的可行性。