ARTIFICIAL INTELLIGENCE AND ITS APPLICABILITY TO GEOGRAPHICAL PROBLEM SOLVING

Artificial Intelligence (AI) may be regarded as an attempt to understand the processes of perception and reasoning that underlie successful problem-solving and to incorporate the results of this research in effective computer programs. At present, AI is largely a collection of sophisticated programming techniques. Many of these techniques are based on the premise that the manner in which knowledge is acquired, organized, accessed and modified in both humans and machines provides the basis of “intelligent” decision-making. The techniques of AI are applicable to a wide variety of geographical problems, including the modeling of individual and aggregate decision-making, and the construction of expert systems and “intelligent” geographical information systems.     

人工冻结梁构件中不同加筋作用的对比试验研究

考虑土层冻结构件中冻结管的加筋作用,对于增加冻结构件强度,减小变形量,提高人工冻结支护工程经济效益有重要作用。为研究冻结管对人工冻结构件的加筋作用,通过淤泥质粘土人工冻结梁构件三点弯试验,获得了各级载荷下构件的位移值。根据对冻土梁加筋与否,构件特定力学响应的比较,完成了冻结管加筋作用的量化分析。     

双星定位系统在中低轨卫星定轨中的应用

对双星定位系统在中低轨卫星轨道确定中的应用做了详细的分析，给出了在该系统的测量模式中低轨卫星定轨的几种方法。通过对这些方法进行的大量模拟计算和分析，结果表明作为双星定位系统的一种潜在功能，用其对中低轨卫星定轨是可行的。     

地质领域机器学习、深度学习及实现语言

地质大数据正在以指数形式增长。只有发展智能数据处理方法才有可能追上大数据的超常增长。机器学习是人工智能的核心,是使计算机具有智能的根本途径。机器学习已成为地质大数据研究的前沿热点,它将让地质大数据插上翅膀,并因此改变地质。机器学习是一个源于数据的模型的训练过程,最终给出一个面向某种性能度量的决策。深度学习是机器学习研究中的一个重要子类,它通过构建具有很多隐层的机器学习模型和海量的训练数据,来学习更有用的特征,从而最终提升分类或预测的准确性。卷积神经网络算法是最为常用的一种深度学习算法之一,它广泛用于图像识别和语音分析等。Python语言在科学领域的地位占据着越来越重要。其下的Scikit-Learn是一个机器学习相关的库,提供有数据预处理、分类、回归、聚类、预测、模型分析等算法。Keras是一个基于Theano/Tensorflow的深度学习库,可以应用来搭建简洁的人工神经网络。     

凡纳对虾人工繁殖的初步研究

研究了海水比重的变化对凡纳对虾亲虾人工繁殖活动的影响及换水对亲虾交配率的影响。试验采用剪眼柄、强化培育、控光、换水等技术措施 ,将凡纳对虾培育成熟并使之自然交配产卵 ,孵化出无节幼体 ;亲虾的平均产卵量可达 1 6万粒左右 ,卵的受精孵化率可达 80 %以上。结果表明 :适时换水可将亲虾的交配率从 4 6 %提高到 79% ;凡纳对虾人工繁殖的适宜海水相对密度为 1 0 1 6以上。     

Mapping of hard rock aquifer system and artificial recharge zonation through remote sensing and GIS approach in parts of Perambalur District of Tamil Nadu,India

Mapping of hard rock aquifer system and artificial recharge zonation were carried out in an area of 325 km2 in parts of the Perambalur District, Tamil Nadu, India. This district has been declared as one of the over-exploited regions in Tamil Nadu by the Central Groundwater Board. To raise the groundwater level, suitable recharge zones were identified and artificial recharge structures are suggested using geomatics technology in the present study. To this end, various thematic maps concerning lithology, soil, geomorphology, land use, land cover, slope, lineament, lineament density, drainage, drainage density and groundwater depth level were prepared. Fissile hornblende gneiss (244 km2) covered most of the study area followed by charnockites (68 km2). Structural hills and rocky pediments characterize the major geomorphological features in the targeted area, and are followed by deep moderated pediments. The area is mostly used as crop and fallow land, followed by scrub land and deciduous forest. In the study area, the slopes are predominantly very gentle (142 km2) and nearly level (66 km2) ones. Besides, Groundwater level data of 58 wells have been generated, in which the minimum and maximum depth were 3 and 28 m respectively. Integration under the GIS environment has been carried out using all the thematic layers to identify the groundwater prospect zone through the introduction of weight and rank methods. Integrated output performances were classified into very poor, poor, moderate, good and excellent categories. All these classes were further divided into two groups as suitable and non-suitable area for the selection of recharge sites. Hard rock fractures were mapped as lineaments from satellite images, and besides that, rose diagram was also generated to find out the trend of the fracture. Furthermore, fracture data of 146 numbers have been collected using Brunton compass to generate rose diagram and were correlated with the rose diagram derived from lineaments. The present study significantly brought up a few areas such as Ammapalayam, Melapuliyur, Senjeri and around Siruvachur for artificial recharge.     

基于HYDRUS-2D的渗滤池回灌地下水数值模拟研究

人工回灌是提高水资源利用水平的重要工程措施。深入研究渗滤池回灌过程机理,基于Hydrus-2D,结合具体渗滤池回灌算例,建立渗滤池回灌的二维饱和-非饱和渗流数值模型。模拟结果表明：回灌水入渗形成的＂湿润峰＂逐渐向饱和带移动,同时饱和带的浸润线逐渐上升,形成不断扩大的饱和区,直到形成一个连续的饱和带。回灌水出流速率经过了迅速增大到一峰值后快速减小,然后又缓慢变小的过程,回灌水入渗速率随时间变化的这一特征与Green-Ampt入渗模型的结论是一致的。     

宁芜盆地白象山矿区人工神经网络 三维成矿定位预测研究

本文基于三维地质环境,综合白象山矿区积累的地质资料和物探成果,首先开展三维地质建模工作,详细刻画了白象山矿区的三维地质结构;在三维地质模型基础上,利用三维空间分析手段对三维控矿因素进行定量挖掘,提取了多种三维控矿因素;最后采用人工神经网络方法进行三维成矿定位预测。预测结果显示,人工神经网络三维成矿定位预测能很好的定位出已知矿体,同时显示,在已知矿体北部及东部的深边部具有较高的成矿概率,可作为开展进一步找矿勘探的靶区。因此,人工神经网络三维成矿定位预测对于白象山矿区的应用是有效的,可服务于新老矿区的深边部三维成矿定位预测,同时可为隐伏矿、盲矿的成矿预测和优选靶区提供定量、定位新的方法和途径。     

桥基隧道明挖施工多元信息集成智能方法与应用

针对大连地铁香工街车站明挖施工邻接桥桩的问题,构建施工远程信息自动化监测体系,进行静力水准仪、电子测斜仪、电子轴力计等传感器布置,获得施工过程多元信息。基于Visual C#.Net平台自主开发了集成智能分析系统,引入支持向量机和差异进化算法,建立三维力学参数反分析和时间序列预测方法。该软件系统包括数据操作、反馈分析、数据报警和可视化输出的模块。将该方法系统应用于大连地铁香工街车站施工过程,进行多元信息采集,利用实际监测的信息数据,调用系统差异进化算法进行第2层和第3层岩石弹性参数搜索,反馈计算值与监测值的最大相对误差为8.23%,进而简化了倒换撑方案。然后通过对QCJ-13-01桥墩监测数据的DE-SVM非线性时间序列预测和超前报警,及时采取隔离桩措施,有效地控制了桥桩变形。该方法系统集自动化远程监测硬件技术和计算智能分析功能于一体,为类似复杂地下工程信息化施工提供先进有效的途径。     

人工智能技术气候预测应用简介

近年来,随着人工智能技术在多个领域大数据分析中的应用,许多研究工作者尝试将地学研究与人工智能跨学科结合,取得了很多新的进展,推动了地球科学的发展。其中气候预测与人类生活以及防灾减灾等息息相关,准确的气候预测至关重要。本文简要总结了人工智能技术在气候预测应用方面的研究进展,包括资料同化、模式参数化、求解偏微分方程、构建统计预测模型、改进数值模式产品释用等领域。这些研究证明了利用人工智能提高气候预测技巧的可能性和适用性,可以极大地节省计算成本和时间。然而人工智能应用也存在诸多挑战,例如数据集的构建、模型的适用性和物理可解释性等问题,对这些难点问题的研究和攻克,可以让人工智能在大数据时代中更好地补充传统地球科学方法,产生更多有益的效应,极大地改进气候预测水平。