波阻抗遗传算法反演方法的研究及应用

作者介绍的遗传算法是在自然界生物遗传过程优选和进化原理的基础上，针对波阻抗反演的实际问题，对标准的遗传算法进行了改进，引入了模拟退火算法中的热槽法控制初始群体的产生，用逐位迭代的交换方式代替传统的杂交方式即随机选择杂交点进行交换。使用改进的遗传算法对地震波阻抗进行反演，可很好地提高反演的精度。在文中最后，作者通过理论模型和实例验证了上述的思想。     

A GIS-based method for flooded area calculation and damage evaluation

Using geographic information system to study flooded area and damage evaluation has been a hotspot in environmental disaster research for years. In this paper, a model for flooded area calculation and damage evaluation is presented. Flooding is divided into two types: ‘source flood’ and ‘non-source flood’. The source-flood area calculation is based on seed spread algorithm. The flood damage evaluation is calculated by overlaying the flooded area range with thematic maps and relating the result to other social and economic data. To raise the operational efficiency of the model, a skipping approach is used to speed seed spread algorithm and all thematic maps are converted to raster format before overlay analysis. The accuracy of flooded area calculation and damage evaluation is mainly dependent upon the resolution and precision of the digital elevation model (DEM) data, upon the accuracy of registering all raster layers, and upon the quality of economic information. This model has been successfully used in the Zhejiang Province Comprehensive Water Management Information System developed by the authors. The applications show that this model is especially useful for most counties of China and other developing countries.     

Incompressible or nearly incompressible soil dynamic behaviour—a new staggered algorithm to circumvent restrictions of mixed formulation

The restriction of the Babuska–Brezzi stability criteria for the mixed formulation requires non-uniform interpolation of displacement and pore-pressure variables, which complicates the coding in the finite element analysis. This will certainly preclude the use of many otherwise useful elements. A possible solution to overcome this problem introduced earlier by the authors requires a semi-explicit form. However, the effectiveness of the semi-explicit procedure is dependent on the time step length used, which presents difficulties for transient analysis. In this paper, an unconditionally stable staggered implicit-implicit algorithm is developed, which is generalised from the semi-explicit form. For non-uniform mesh configurations, the results indicate a significant improvement compared with those obtained by using the semi-explicit procedure.     

强震前中期地震活动的变化及TIP预测研究

用改进的Ｍ８算法研究了１９７９年以来我国华北西南及邻近地区共１６次强震前的ＴＩＰ，即震发生概率增长时间。结果表明，１４次强震发生在补判定为概率增长时间的ＴＩＰ内。ＴＩＰ警戒约占研究部时空域的３７％，获得了较好的强震中期预测内符效果。表明该方法可作为强震中期预测的手段之一。     

Automatic identification of rocks in thin sections using texture analysis

This paper presents a test of the automatic identification of photomicrographs of rocks in thin sections using digital image processing and texture analysis. Three sets of textural measures derived respectively from the cooccurrence matrix, texture space, and texture spectrum have been used to identify six rock types (mylonite, diorite porphyry, diorite. gabbro, granite, and peridotite). Exploring only the texture characteristics of images, the average correct recognition rate reaches 89% for 58 photographs belonging to the six types. The results also show the importance of using those more discriminating texture features in the classification algorithm. For the present study, the average rate of correct classification ranges from 46% to 89% depending on the set of texture measures used. The features extracted from the texture spectrum have more discriminating performance than the conventionally used Haralick measures derived from the cooccurrence matrix.     

平面控制网网点优化排序研究

本文利用法方程系数阵的稀疏性,将Banker算法和RCM算法应用于平面控制网数据处理,实现网点自动优化排序,压缩了法方程系数阵的存贮空间。借助于各类平面网平差模型对Banker 算法和RCM算法进行了比较、分析。采用优化排序算法不仅能扩大微机的解算容量,同时也提高了解算速度,减少了计算过程中的舍入误差,且适用于任意平面网、高程网,GPS网的自动化数据处理。     

四川地震C值扫描：算法,格式,数据结论

为检验地震C值异常是否可以作为大中地震的前兆指标,就需要在两者之间进行大规模的比较性研究。因此,至少应在现代地震台网的时空控制范围内,对C值进行扫描。本文的算法体系是以极大似然法加统一格式加时间窗滑动构成的。这种选择,既有在技术上对前人工作的继承与发展,也反映了笔者面向唯象性实用的价值观念。至于效果如何,仅以四川全省C值扫描为例。     

前馈神经网络在矿物识别方面的应用

本文首先介绍人工神经网络的发展现状，然后着重阐述BP人工网络的基本原理和算法，并以蓝宝石的产地识别和石榴子石分类识别作为具体实例，用此说明人工神经网络方法在矿物学中的应用，最后对识别结果及BP算法本身进行了讨论。     

四叉树结构在数字图像分割中的应用

本文讨论了四叉树的基本概念、图像的四叉树表示以及四叉树结构在计算机中的实现；给出了用这种算法实现分裂与合并算法的步骤。实验表明，把四叉树结构引入图像分割，收到了较好的效果。     

倾斜相机式投影的几何解法及其与外心投影的比较

本文根据透视投影的基本原理，给出了关于倾斜相机式投影的几何解法，在此基础上分析了倾斜相机式投影与外心投影的异同，提出了二者比较的一致性条件，并给出了实例验证。