基于光谱和纹理的SVM矿化蚀变信息提取研究

针对传统矿化信息提取方法单一,利用光谱或纹理、信息量相对较少、需要大量样本的缺陷,利用基于光谱和纹理的支持向量机(SVM)原理,建立矿化信息提取模型.选择青海泽库县析界日地区作为典型研究区.首先提取研究区光谱和纹理信息,选取训练样本;然后求解最优超平面,进而确定决策函数;最后泛化推广识别其他待识别的样本.通过所提取的遥感蚀变异常信息与重砂异常点叠加分析,叠加基本吻合;从野外实地验证来看,均发现了不同程度的矿化现象,并指出了5个重点异常区.     

Geologic Characteristics of Gas Reservoirs in West Sichuan Foreland Basin

The foreland basin in West Sichuan is a tectonic unit that has undergone multi-periods tectonic movements of Indosinian-Yanshanian-Himalayan. Since late Triassic, it has been in a passive subsidence environment controlled by basin margin mountain systems and by the compression with abundant sediment sources. With the complex geologic setting, the main geologic characteristics of natural gas reservoir are listed as following:(1)Source rocks are coal-bearing mud and shale series with high to over maturity, and long and progressive hydrocarbon generation-displacement period. The key accumulation period is middle-late Yanshanian epoch.(2)There are three gas-bearing systems vertically, each of which has different reservoir mechanism, main-controlled factors and distribution law, so the exploration thoughts and techniques are also different.(3)Undergoing multi-period generation-migration-accumulation, oil and gas have encountered multi-period modification or destruction, and gas accumulation overpass multiple tectonic periods. So the trap type is complicated and dominated by combination traps. Because the main accumulation period of natural gas is early and the reservoir encountered the modification of strong Himalayan movement, there is great difference in the fullness degree of gas reservoirs and complicated gas-water relation. (4) Reservoir is tight to very tight, but reservoirs of relatively high quality developed under the super tight setting. (5) The key techniques for oil and gas exploration in west Sichuan foreland basin are the prediction of relatively favorable reservoirs, fractures and gas bearing; and the key techniques for oil and gas development are how to improve the penetration rate, reservoir protection and modification.     

RBF网络在地质样品非线性基体效应校正中的应用

基于RBF(Radial Basis Function)网络的正交最小平方算法,对攀枝花地质样品进行动态非线性基体效应校正.首先采用EDXRF仪器对样品进行测量,并对样品的荧光计数进行归一化,设计出自动分类模型和动态非线性数据基体效应校正模型的二级串联结构,用自组织神经网络(SOFM)进行分类,然后用RBF网络预测了未知样品元素Ti的百分含量.结果表明:对所预测样本的Ti的百分含量与化学分析相比的误差均小于0.5%.由此可见,用RBF方法可以有效地对地质样品进行非线性基体效应校正,能够达到工矿生产的要求.     

基于约束的城市街道网自动综合方法

近年来,基于约束的地图自动综合成为研究的热点。本研究在基于约束的地图自动综合理论的基础上,提出了一种城市街道网自动综合方法。该方法依赖于街道网综合中的约束和改进的动态决策树结构,它很好地将约束融入了特定的数据结构,实现了街道网的渐进式综合。实验证明了方法具备可行性。     

重庆市石漠化灾害特征及防治分区研究

石漠化灾害是影响重庆市岩溶地区可持续发展的主要障碍之一,必将危及到三峡水库区的生态环境.在遥感和GIs技术的支持下,文章对重庆市石漠化灾害的数量特征、空间分布特征,以及石漠化灾害的发育特征进行分析,进而对研究区的石漠化灾害治理进行分区,并对每一分区的石漠化灾害的特征与成因进行分析与描述,提出石漠化灾害治理与防治的主要措施.     

含盐量对松嫩平原碳酸盐渍土抗剪强度的影响

松嫩平原分布着大量的碳酸盐渍土,其不良的工程性质给穿越其中的道路质量造成不同程度的影响。由于盐渍土所含易溶盐成分主要为碳酸氢钠,该研究通过向碳酸盐渍土内加入碳酸氢钠制备不同含盐量的盐渍土,进行室内直剪试验。试验表明：随着含盐量的增加土的粘聚力减小,内摩擦角有先减小后变大的趋势;随着含水量的增加,土的含盐量越高,土的粘聚力减小越明显。易溶盐在土中的状态形式是影响其抗剪强度参数变化的一个重要原因。     

丽江市地质灾害易发性遥感影像判译

丽江市地质灾害多发,在滇西北旅游城市中较典型。应用遥感技术对丽江市地质灾害易发性的遥感研究,可突破传统方法,发挥宏观、综合、直观、快速特点,取得较好效果。