【作者】 陈聆；

【导师】 郭科；

【作者基本信息】 成都理工大学 ， 地球探测与信息技术， 2011， 博士

【摘要】 本论文选题来源于中国地质调查局地质调查工作项目“火山岩盖层下找矿技术攻关研究”子课题:地球化学矿致异常定量评价方法研究。由于火山岩盖层对矿体的封闭作用,对成矿元素的向上扩散和微渗透作用具有很强的阻滞能力,使传统地球化学异常的评价解释变得非常困难。要突破传统,就需要应用新理论、新方法,除了要加强深穿透地球化学等新的化探技术方法研究以外,从数学地质和地质统计学角度,应用现代非线性科学的最新成果,对化探异常的再评价,再解释,进一步挖掘化探数据中蕴含的成矿信息无疑也是一条重要的途径。本论文研究区是西藏洞嘎普铜矿勘查区,论文从地球化学矿致异常定量评价方法研究出发,利用多种现代数学方法和非线性分析方法,挖掘化探数据蕴含的成矿弱异常信息。首先利用人工免疫算法构建了地球化学数据降噪模型,对地球化学元素数据进行降噪处理,以此消除特高值带来的屏蔽及剔除假异常;其次,根据盲源分离算法建立反演化探数据元素组合模型,以此确定地球化学成矿元素组合;然后,利用分形含量面积法确定地球化学单元素及元素组合的异常下限,再利用分形含量梯度法圈定异常浓集中心,进而确定异常分带性;最后,将元素分带特性研究与研究区地质特征相结合,对比单元素异常图及组合异常图,对研究区的地球化学元素作出异常分类和异常评价解释。本论文的研究将为火山岩盖层下的化探找矿技术方面提供新的技术方法。本文的创新点在于:(1)论文在地球化学异常识别方面采用了多种非线性分析方法,克服了传统方法中将地球化学数据看成是符合正态分布或对数正态分布而带来的不合理性,以及传统上将地球化学元素分布曲面看成是连续光滑曲面的不科学性。(2)采用人工免疫理论对地球化学原始数据进行降噪处理,剔除了特高值带来的屏蔽和假异常,使数据更具有效性。(3)由于地球化学元素分布曲面是分形曲面,因而本论文通过分析多种分形方法,使用了将陈秋明的含量-面积模式及谢和平的投影覆盖法结合并改进的分形含量面积法来确定异常下限,并用分形含量梯度法确定异常浓集中心,以此圈定异常分带。(4)论文首次尝试将盲源分离技术用于地球化学领域中,即将盲源分离理论中的FastICA算法应用于地球化学数据寻求区域地球化学矿化元素组合之中,形成一套表生地球化学自动求取元素组合的方法模型,从而反演出地球化学各元素组合特征。(5)论文开展了非线性科学与高新信息处理技术相结合的地球化学数据处理新方法的研究,为多尺度地球化学异常圈定及深部成矿预测理论提供了新的思路,研究选题不仅具有新颖性和实用性,而且所提出的具有扎实工作基础的研究方案将为现有的化探工作提供有益补充。通过研究,在MAPGIS下建立了地球化学异常圈定系统,对西藏自治区洞噶普铜矿的1:1万土壤地球化学测量资料进行处理后,绘制了测区地球化学元素的异常图和组合异常图,并进行了异常解释评价。经过验证于测区内共划分出4个异常,其中甲(1)类异常2个,编号为Ht-甲(1)-1,Ht-甲(1)-2;甲(2)类异常1个,编号为Ht-甲(2)-1;乙类异常1个,编号为Ht-乙-1。所划的异常整体分布反映的元素套叠状况、异常强度均反映了测区的实际状况。异常与目前发现的矿体、矿化体吻合。异常完整、闭合好,表明异常属于矿致异常。其中Ht-甲(1)-1异常与Ⅲ号强矿化带吻合,Ht-甲(1)-2异常与Ⅳ号矿体吻合,Ht-甲(2)-1异常位于③号韧性剪切带上盘,与2004年发现的Ⅴ号矿体关系密切。更多还原

【Abstract】 This article is from the Geological Survey projects of China Geological Survey, sub-topics of‘the exploration technology research Under the volcanic cover’: quantitative evaluation method of Geochemical ore-forming anomaly. Since the closure of volcanic cover is that the body effect on the forming elements of the upward diffusion and micro-infiltration block has a strong ability to evaluate the traditional interpretation of geochemical anomalies become very difficult. To break through the traditional, you need to apply new theories, new methods. n addition to strengthening the deep-penetrating geochemical exploration of new technology and methods of research other than geology and geostatistics from the mathematical point of view, the application of the latest achievements of modern nonlinear science, the re-evaluation of geochemical anomalies, to explain, and further excavation geochemical metallogenic information contained in the data is undoubtedly an important way.This study area is the Dong Gapu copper exploration area in Tibet. From quantitative evaluation method of Geochemical ore-forming anomaly, making use of a variety of modern mathematical methods and nonlinear analysis, mining ore weak anomalies contains in geochemical data. First use of artificial immune algorithm to construct a noise model of geochemical data, geochemical elements data on noise reduction, in order to eliminate high-value brought by special shielding and removing false anomaly. Secondly, based on blind source separation algorithm for establishing anti-evolution geochemical data elements combined model, in order to determine combination of geochemical ore-forming elements. Then, using the fractal concentration-area method to determine single-element geochemistry and elemental composition of the anomaly threshold, re-use content gradient method of fractal delineate abnormal concentration center and to determine the zoning exception. Finally, the elemental characteristics of research and study area with geological characteristics of the combination, compared to single-element anomaly maps and the composite anomaly map of the study area of the geochemical elements to explain the anomaly classification and anomaly evaluation. This study will provide new technical methods for geochemical exploration technical aspects under the volcanic cover.Innovation of this article is:(1)The article in terms of geochemical anomaly identified by a variety of nonlinear analysis method to overcome the traditional methods of geochemical data will be seen as consistent with normal or lognormal distribution brought about irrationality, and the distribution of surface geochemical elements traditionally seen as a continuous smooth surface of the unscientific nature.(2)Artificial immune theory of noise on the geochemistry of the raw data processing, removing the special shielding to bring high-value and false exceptions, to make the data more effective.(3) The surface geochemical element distribution is fractal surface, and thus this article, through analysis of various fractal methods, the use of the content-area model of Chen Qiuming and the Projective covering method of Xie Heping and improved method combined with the fractal method to determine anomaly threshold, and the gradient method used to determine the fractal content of abnormal concentration center, as delineated anomalous zone.(4) The article first attempt to blind source separation technique for geochemical field, blind source separation theory about the FastICA algorithm is applied to regional geochemical data for mineral elements in combination, forming a supergene geochemical method of automatic model to strike a combination of elements, which combined the elements of anti-geochemical characteristics.(5) The article carried a nonlinear information processing science and high technology combined geochemical data processing method of research for multi-scale delineation of geochemical anomalies and deep metallogenic prediction theory provides a new way of thinking, research topics not only novel and practical, but also has a solid work proposed research program based on the existing geochemical exploration will provide a useful supplement.Through research, established geochemical anomaly delineated system under the MAPGIS . after the data processing of 1:1 million soil geochemical survey of Dong Gapu copper in Tibet Autonomous Region, drawing elements of the surveyed area geochemical anomaly maps and the combination of anomalies , and had anomaly interpretation and evaluation. Validated in the survey area was divided into four abnormal. There have two A (1) class anomaly , numbers are Ht-A (1) -1 and Ht-A (1) -2; one is A (2) class anomaly , number is Ht-A (2) -1; one is anomaly B , number is Ht-B -1. The overall distribution of anomalies by overlaying the elements reflected in the situation, unusual intensity reflects the actual situation of the surveyed area. Anomalies and ore bodies found so far, consistent with mineralization. anomaly complete, close is good, that anomaly is an anomaly of mineralization. Where Ht-A (1) -1 andⅢof ore-forming anomaly with strong fit, Ht-A (1) -2 anomaly consistent withⅣorebody, Ht-A (2) No -1 anomaly in the ductile shear belt drive③, found in 2004 closeⅤorebody.更多还原