利用目标点最小生成树的路面裂缝检测

分析路面裂缝的成像特点,提出了一种基于目标点最小生成树的路面裂缝检测方法。最小生成树具有描述空间线性聚集特征的优点,有利于实现对裂缝的全局分析和提取。实验结果表明,对不连续裂缝的处理,本文方法优于传统方法。     

海量矿业权核查成果数据存储与访问的优化技术研究

针对矿业权核查单矿权成果数据具有文件小、数量多、异构、多源等特点,为了有效地存储这些数据到Oracle数据库中,实现快速提取及显示,并能够将一个矿业权的提取时长控制在1秒以内,本文给出了从建立数据库实体到查询、显示过程所涉及诸多问题的一个完整的优化方案。首先建立了一个完整的、科学的数据存储模型,并采用范围-列表分区模式对大型表进行分区,同时采用通过调整Oracle系统高速缓冲区、减少表连接、优化SQL查询语句、在客户端分页显示查询结果等优化方案,有效提升表中数据的查询速度,以达到优化目的。     

利用决策树工具的土地利用类型遥感识别方法研究

应用决策树的理论和方法,利用遥感数据及其他相关数据和资料进行土地利用信息分类。通过研究地物光谱统计特征,讨论了通过耕地指数等归一化地类指数来增强影像地类特征、结合DEM提取土地利用信息的决策树分支点的设计方法,较好地解决了水体和建筑阴影、道路等容易混淆区域的区分问题。     

相空间重构在滑坡预测中的应用

针对滑坡位移时间序列的非线性特性,引入基于相空间重构和最小二乘支持向量机(LSSVM)的预测法.利用Cao氏方法确定嵌入维数,根据互信息法计算最佳延迟时间;然后在相空间中,利用最小二乘支持向量机(LSSVM)建立预测模型,对滑坡进行了实证计算,且与LSSVM模型和BP神经网络模型进行了比较.结果表明,模型具有较高的精度...     

一种利用K均值和SOM进行遥感图像分类的方法

探讨了一种将K均值算法和SOM神经网络算法相结合的方法,并将其应用于多光谱遥感图像分类,通过与K均值算法、ISODATA算法和SOM算法的对比实验,验证了该方法的有效性.     

平坦地区GPS高程异常拟合研究

GPS测量所提供的高程为相对于WGS-84椭球的GPS大地高,而我国使用的是正常高。大地高等于正常高与高程异常之和,要使GPS高程在工程实际中得到应用,必须先求出高程异常,进而获得正常高。结合GPS测量和水准测量资料,用神经网络方法和二次多项式曲面拟合方法拟合高程异常,对拟合精度进行了分析比较,得出了有实用价值的结论。     

基于MATLAB神经网络方法的多层砖房震害预测

提出利用MATLAB人工神经网络工具箱建立基于贝叶斯正则算法的BP神经网络模型,以地震区多层砖房震害调查数据为因子的震害预测方法.神经网络模型输入震害因子包括建筑的层数、施工质量、房屋整体性等,输出值为建筑物在地震作用下的破坏程度.结果表明,本方法可以对多层砖房的震害样本进行预测并达到较理想的效果.     

Bioturbation on a south‐east Australian hillslope: estimating contributions to soil flux

The contribution of bioturbation to downslope soil transport is significant in many situations, particularly in the context of soil formation, erosion and creep. This study explored the direct flux of soil caused by Aphaenogaster ant mounding, vertebrate scraping and tree‐throw on a wildfire‐affected hillslope in south‐east Australia. This included the development of methods previously applied to Californian gopher bioturbation, and an evaluation of methods for estimating the volume of soil displaced by tree‐throw events. All three bioturbation types resulted in a net downslope flux, but any influence of hillslope angle on flux rates appeared to be overshadowed by environmental controls over the spatial extent of bioturbation. As a result, the highest flux rates occurred on the footslope and lower slope. The overall contribution of vertebrate scraping (57.0 ± 89.4 g m^?1 yr^?1) exceeded that of ant mounding (36.4 ± 66.0 g m^?1 yr^?1), although mean rates were subject to considerable uncertainty. Tree‐throw events, which individually cause major disturbance, were limited in their importance by their scarcity relative to faunalturbation. However, tree‐throw might be the dominant mechanism of biotic soil flux on the mid‐slope provided that it occurs at a frequency of at least 2–3 events ha^?1 yr^?1. Although direct biotic soil flux appears to be geomorphologically significant on this hillslope, such transport processes are probably subordinate to other impacts of bioturbation at this site such as the enhancement of infiltration following wildfire. Copyright © 2011 John Wiley & Sons, Ltd.     

Artificial neural network forward modelling and inversion of electrokinetic logging data

An artificial neural network method is proposed as a computationally economic alternative to numerical simulation by the Biot theory for predicting borehole seismoelectric measurements given a set of formation properties. Borehole seismoelectric measurements are simulated using a finite element forward model, which solves the Biot equations together with an equation for the streaming potential. The results show that the neural network method successfully predicts the streaming potentials at each detector, even when the input pressures are contaminated with 10% Gaussian noise. A fast inversion methodology is subsequently developed in order to predict subsurface material properties such as porosity and permeability from streaming potential measurements. The predicted permeability and porosity results indicate that the method predictions are more accurate for the permeability predictions, with the inverted permeabilities being in excellent agreement with the actual permeabilities. This approach was finally verified by using data from a field experiment. The predicted permeability results seem to predict the basic trends in permeabilities from a packer test. As expected from synthetic results, the predicted porosity is less accurate. Investigations are also carried out to predict the zeta potential. The predicted zeta potentials are in agreement with values obtained through experimental self potential measurements.     

集对分析径向基函数神经网络预测模型

将集对分析与径向基函数神经网络结合,提出了集对分析径向基函数神经网络预测模型。模型思路是将研究对象t-1时和t时的影响因子集构造为集对并计算联系度,由联系度的同一度、差异度、对立度及研究对象t-1时的值为输入,研究对象t时的值为输出,构建径向基函数神经网络。以年径流预测为例研究表明,模型结构清晰、步骤明确、预测精度较高,为集对分析应用于水文预测提供了新思路。