利用人工神经网络自动识别煤田测井岩性

介绍采用人工神经网络（ＡＮＮ）模型，借助于误差逆转播算法，应用到煤田测井岩性自动识别中，效果较好。为提高该方法的实用性，通过对误差逆传播算法的改进，并经过验算，表明了其优越性；文中采用多层人工神经ＢＰ网络模型，对较大样本（４８组）进行学习，可以识别８种岩性，说明了该方法的实用性。     

基于异步迭代算法的冲击地压预测

采用三层BP神经网络方法对冲击地压建立了数学模型。网络的训练算法采用具有松弛因子的动态异步迭代法,该算法在克服网络的麻痹现象及局部极小问题上都优于目前常用的训练方法,因而,采用此算法对网络进行了训练及震级的预报。基于新汶矿务局华丰煤矿1999～2000年的冲击地压现场监测数据,把冲击地压的能量、产生波的幅值、频次做为输入数据,相应期间的最大震级为输出数据,组成神经网络的训练样本及测试样本,对原始数据进行了数学预处理,网络结构采用了输入层3个结点,中间层7个结点,输出层1个结点的前向神经网络;网络最终的训练误差为0.06,预测结果的相对误差率平均为 9.2 %,预测效果比较理想。     

福建德化西洋坂金矿地质特征及找矿前景

西洋坂金矿16个矿(化)体均产于其矿源层葛坑组变质岩中,属变质热液型,根据地层、构造、土壤地球化学异常分析,区内北东、北西向断裂交汇处及其深部有找到新的金矿前景。     

扬子地块西缘震旦系灯影组层控铅锌矿床的成矿规律及形成超大型矿床的可能性

扬子地块西缘灯影组层控铅锌矿床是该区最重要的一类矿床。成带分布，受构造控制，具明显的后生特征。所有矿体均产在灯影组碳酸盐建造顶部，并靠近下寒武统砂页岩建造。受岩相组合（地球化学障）的控制，属多源混合后生热液成因层控矿床。根据矿床成因和成矿规律的研究，本区灯影组具备形成超大型矿床的潜力。     

扬子地块西缘震旦系灯影组层控铅锌矿床的成矿规律及形成超大型矿床 …

扬子地块西缘灯影组层控铅锌矿床是该区最重要的一类矿床。成带分布，受构造控制，具明显的后生特征。所有矿体均产在灯影组碳酸盐建造顶部，并靠近下寒武统砂页岩建造、受岩相组合（地球化学障）的控制，属多源混合后生热液成因层控矿床。根据矿床成因和成矿规律的研究，本区灯影组具备形成超大型矿床的潜力。     

利用神经网络技术在储层非均质性上的研究

储层非均质性对油气田的勘探和开发效果影响很大。结合测井资料二次解释的孔隙度(φ)、泥质含量(Vsh)、粒度中值(Md)、束缚水饱和度(Swi)等资料,借助BP网络算法,对渗透率进行预测。BP网络算法是人工神经网络算法中的一种,采用的是误差逆传算法,即把网络输出的误差归纳为各连接权的过程,通过把输出层的误差逐层向输入层逆向传播以分摊给各层单元,从而获取各层单元的参考误差,以便调整相应的联接权。如此反复,直至网络输出与期望输出之间的误差减少到允许误差为止。通过这样学习过程,消除了系统误差,提高了渗透率预测的精度,并为定性定量描述储层非均质性打下了坚实的基础。通过对萨尔图油田中部葡萄花油层组的非均质性进行了定量半定量的研究,认为萨尔图油田中部葡萄花油层组PI2小层层内非均质性最严重,与沉积微相密切相关。单砂层垂向上渗透率的变化以正韵律和复合式韵律为主,局部发育反韵律模式。经对葡萄花油层组PI1~PI4系统研究,层内非均质性强弱依次为PI2a、PI3、PI2b、PI1、PI4;层间非均质性在PI3和PI4间表现得最强,其他相对要弱;各小层平面非均质性相差无几,整体表现为较强的非均质性。     

系统滤波模型在滇池水质预测中的应用

根据滇池水深较浅，流场比较稳定等特点，对湖泊进行了单元划分，考虑计算误差和观测误差干扰的存在，建立了系统滤波模型，用吉尔算法求解微分方程，对ＢＯＤ，ＣＯＤ多点连续模拟预测，经用１９８８年实测资料检验，取得满意结果。     

复杂表层地区地震资料处理中的静校正

利用西北某区资料，探讨在复杂表层地震地质条件下，如何做好浅目的层、短排列接好地震资料处理中的静校正工作，以ＧＥＯＶＥＣＴＥＵＲＰＬＵＳ处理软件为基础，对野外静校正参数进行误差分析，根据误差大小和分布规律，用不同方法分阶段对其修正，确保了最终自动剩余静校正前，静校正误差减小到目的层反射波半个周期以内。     

海流图盆地铀矿化（异常）成因及找矿前景

海流图盆地内铀矿化(异常)均产出于西部李三沟组中段砂岩层中，分析认为具有层间氧化带砂岩型铀矿化特征。本文以盆地地质背景、铀源条件为基础，着重分析了盆地内铀矿化(异常)的成因和砂岩型铀成矿条件，认为盆地西部存在有利的铀矿成矿环境及找矿远景。     

云南宁蒗地区铜厂河铜矿构造控矿规律研究

宁蒗地区铜厂河铜矿主干控矿构造为SN向断裂体系，起着导矿、配矿、容矿的作用。矿区NWW向F1断裂为阻矿构造，铜矿化均产于其下盘（南盘）；NE向F2，F3断裂将本区分成三个不同矿化特征的矿段，其东、中矿段是寻找隐伏－半隐伏铜矿体的有利地段。由于构造活动的多期次和构造形变的分带性，矿区铜矿具显著的成矿多期、多阶段性和矿化体水平－垂直分带性规律。     

基于神经网络在路面基层压实参数中的应用

石灰工业废渣稳定类半刚性材料是高等级公路路面基层材料常见形式之一,根据规范和设计要求可分为含骨料类和不含骨料类。当骨料含量超过50％时,室内重型击实试验劳动量大,并且干密度和含水率睦线不稳定。在已知不含骨料的石灰工业废渣稳定类半刚性材料（即结合料）的最大干密度和最优含水率的基础上,通过结合人工神经网络理论,基于Matlab的BP人工神经网络,建立并编制了含骨料的石灰工业废渣稳定类半刚性材料压实参数（最大干密度和最优含水率）的预测网络模型,经过对网络模型的大量训练、训练函数和传递函数的调整及初始训练数据的规一化,最后建立了6→15→2的网络模型,其网络模型预测结果稳定准确,有一定实际应用价值。     

地质调查信息网格GIS平台山东结点建设探讨

地质调查信息网格平台是基于网格GIS技术,空间数据库具有开放性、集成性,可以对地质调查信息化成果进行多样式、海量数据的组织集成与分布、网络共享服务。结点是平台系统中实现数据资源无缝集成与共享的中间件,是其核心部分。该文叙述地质调查信息网格平台山东结点建设中应用的技术方法。     

A stochastic optimization approach for paper recycling reverse logistics network design under uncertainty

One of the most important objectives of a manufacturing firm is the efficient design and operation of its supply chain to maximize profit. Paper is an example of a valuable material that can be recycled and recovered. Uncertainty is one of the characteristics of the real world. The methods that cope with uncertainty help researchers get realistic results. In this study, a two-stage stochastic programing model is proposed to determine a long term strategy including optimal facility locations and optimal flow amounts for large scale reverse supply chain network design problem under uncertainty. This network design problem includes optimal recycling and collection center locations and optimal flow amounts between the nodes in the multi-facility environment. Proposed model is suitable for recycling/ manufacturing type of systems in reverse supply chain. All deterministic, stochastic models are mixed-integer programing models and are solved by commercial software GAMS 21.6/CPLEX 9.0.     

The Importance of the Precipitation and the Susceptibility of the Slopes for the Triggering of Landslides Along the Roads

In order to characterise the influence of the heavyrains on the observed landslides during the 1996–1997hydrological cycle, rainfall records for the last 100years are analysed from 104 stations in easternAndalusia. Regarding the amounts of rain recordedbetween October 1996 and March 1997 in the 104stations studied, 31 presented new all-time records;15 presented values that were 80–100% of thepre-1995 record; 49 stations, 80–50%; and 9stations, < 50%. A map has been devised of thesusceptibility of the materials through which thesouth-eastern Andalusian road network crosses,together with an inventory of the damage caused byinstability phenomena on banks and cuttings of theroad network during the winter of 1996–1997. Therelationships between the rainfall during the studyperiod, the damage caused to the road network and thesusceptibility of the materials affected are analysed.The results indicate that there is a clearcorrespondence between the rainfall recorded and thesusceptibility of the materials with the inventorieddamage. It is concluded that the widespread seriousdamage caused in early 1997 to the roads andsurrounding areas in the Alpujarra region and thecoast of the Province of Granada was mainly caused bythe extraordinarily heavy rains. However, considerablyless damage was observed where the susceptibility ofthe terrain is low, thus highlighting the extremeusefulness of terrain-susceptibility maps for riskprevention and territorial planning.     

GPS高程拟合的方式及可靠性分析

在范围不大的区域中，高程异常具有一定的几何相关性，GPS高程拟合就是利用这一原理，求解正常高。在解析法求解过程中，首先用最小二乘法确定拟合数学模型的系数，在此基础上计算出待测点的高程异常值。通过实例验证：GPS高程拟合的精度主要取决于GPS大地高的精度、重合点正常高的精度、重合点的分布及拟模型的选择。一般在重合点数量充足且分布均匀的情况下，GPS高程拟合的精度可达到四等水准网的精度要求。     

Multivariate forecast of winter monsoon rainfall in India using SST anomaly as a predictor: Neurocomputing and statistical approaches

In this article, the complexities in the relationship between rainfall and sea surface temperature (SST) anomalies during the winter monsoon over India were evaluated statistically using scatter plot matrices and autocorrelation functions. Linear, as well as polynomial trend equations were obtained, and it was observed that the coefficient of determination for the linear trend was very low and it remained low even when polynomial trend of degree six was used. An exponential regression equation and an artificial neural network with extensive variable selection were generated to forecast the average winter monsoon rainfall of a given year using the rainfall amounts and the SST anomalies in the winter monsoon months of the previous year as predictors. The regression coefficients for the multiple exponential regression equation were generated using Levenberg-Marquardt algorithm. The artificial neural network was generated in the form of a multilayer perceptron with sigmoid non-linearity and genetic-algorithm based variable selection. Both of the predictive models were judged statistically using the Willmott's index, percentage error of prediction, and prediction yields. The statistical assessment revealed the potential of artificial neural network over exponential regression.     

Hydropyrolysis of insoluble carbonaceous matter in the Murchison meteorite: new insights into its macromolecular structure

The major organic component of carbonaceous chondrites is a solvent-insoluble, high molecular weight macromolecular material that constitutes at least 70% of the total organic content in these meteorites. Analytical pyrolysis is often used to thermally decompose macromolecular organic matter in an inert atmosphere into lower molecular weight fragments that are more amenable to conventional organic analytical techniques. Hydropyrolysis refers to pyrolysis assisted by high hydrogen gas pressures and a dispersed catalytically-active molybdenum sulfide phase. Hydropyrolysis of meteorites has not been attempted previously although it is ideally suited to such studies due to its relatively high yields. Hydropyrolysis of the Murchison macromolecular material successfully releases significant amounts of high molecular weight PAH including phenanthrene, carbazole, fluoranthene, pyrene, chrysene, perylene, benzoperylene and coronene units with varying degrees of alklyation. Analysis of both the products and residue from hydropyrolysis reveals that the meteoritic organic network contains both labile (pyrolysable) and refractory (nonpyrolysable) fractions. Comparisons of hydropyrolysis yields of Murchison macromolecular materials with those from terrestrial coals indicate that the refractory component probably consists of a network dominated by at least five- or six-ring PAH units cross-linked together.     

基于改进原型网络的小样本古生物化石识别研究

传统古生物化石鉴定方法多依赖于古生物学家的经验知识,现有的人工智能识别方法需要大量的化石训练样本才能达到较高的准确率。为解决上述问题,在少量化石图像样本情况下准确识别化石,本文尝试使用残差网络和注意力模块相结合的方法,并将其运用于小样本的化石鉴定。首先以残差网络作为模型的特征提取模块,并在残差网络的残差块中嵌入CBAM卷积注意力模块,提高模型对于化石纹理特征的关注,以提取更为全面的深层次化石图像特征,再使用小样本度量学习中的原型网络对提取特征进行原型计算,最后通过多轮次迭代训练得出最佳的化石判别模型。使用本文方法与5种经典的小样本学习方法进行对比实验,实验结果表明本文方法的识别准确率最高,在样本数量为1和5的情况下,准确率达到了86.32%和94.21%,对稀缺样本下的化石识别具有更显著的优势。     

基于改进原型网络的小样本古生物化石识别研究

传统古生物化石鉴定方法多依赖于古生物学家的经验知识,现有的人工智能识别方法需要大量的化石训练样本才能达到较高的准确率。为解决上述问题,在少量化石图像样本情况下准确识别化石,笔者等尝试使用残差网络和注意力模块相结合的方法,并将其运用于小样本的化石鉴定。首先以残差网络作为模型的特征提取模块,并在残差网络的残差块中嵌入CBAM卷积注意力模块,提高模型对于化石纹理特征的关注,以提取更为全面的深层次化石图像特征,再使用小样本度量学习中的原型网络对提取特征进行原型计算,最后通过多轮次迭代训练得出最佳的化石判别模型。使用本文方法与5种经典的小样本学习方法进行对比实验,实验结果表明本文方法的识别准确率最高,在样本数量为1和5的情况下,准确率达到了86.32%和94.21%,对稀缺样本下的化石识别具有更显著的优势。     

Prediction of longitudinal dispersion coefficient in natural rivers using a cluster-based Bayesian network

The longitudinal dispersion coefficient is a key element in determining the distribution and transmission of pollution, especially when cross-sectional mixing is completed. However, the existing predictive techniques for this purpose exhibit great amounts of uncertainty. The main objective of this study is to present a more accurate model for predicting longitudinal dispersion coefficient in natural rivers and streams. Bayesian network (BN) approach was considered in the modeling procedure. Two forms of input variables including dimensional and dimensionless parameters were examined to find the best model structure. In order to increase the performance of the model, the clustering method as a preprocessing data technique was applied to categorize the data in separate groups with similar characteristics. An expansive data set consisting of 149 field measurements was used for training and testing steps of the developed models. Three performance evaluation criteria were adopted for comparison of the results of the different models. Comparison of the present results with the artificial neural network (ANN) model and also well-known existing equations showed the efficiency of the present model. The performance of dimensionless BN model 30% is more than dimensional ones in terms of the root mean square error. The accuracy criterion was increased from 70 to 83% by performing clustering analysis on the BN model. The BN-cluster model 43% is more accurate than ANN model in terms of the accuracy criterion. The results indicate that the BN-cluster model give 16% better results than the best available considered model in terms of the accuracy criterion. The developed model provides a suitable approach for predicting pollutant transport in natural rivers.