【作者】 王雄军；

【导师】 彭省临；

【作者基本信息】 中南大学 ， 国土资源信息工程， 2008， 博士

【摘要】 云南都龙老君山地区是我国滇东南锡成矿带上最重要的超大型锡、锌多金属矿集区之一,也是中国三大锡矿基地之一,铟矿储量更是全国第一,闻名中外。论文以地洼学说多因复成递进成矿理论和成矿预测理论为指导,以老君山矿集区内不同类型矿床（体）成矿地质地球化学特征、变质岩原岩恢复、矿床成矿作用以及遥感信息解译等方面的研究为切入点,结合区域成矿背景分析,总结了老君山矿集区成矿地质条件、微量元素分布规律、构造演化特征、成矿序列以及成矿演化模式,并采用多因复成递进成矿理论、分形理论、BP神经网络评价模型和矿体空间信息成矿预测模型等方法开展了老君山矿集区控矿机理、遥感信息解译、地质建模及空间信息成矿预测研究。论文的研究成果与创新认识主要体现在以下几个方面:1、通过对矿集区内各种构造交汇格局、构造形迹及其配套构造发育特征的分析,全面系统地阐明了不同构造演化阶段及构造体系的发展、演化特征,指出老君山穹隆、文山—麻栗坡断裂及马关—都龙断裂对成矿具有重要的控制作用。2、在因子分析、聚类分析的基础上,应用多重分形方法对老君山矿集区17个微量元素进行分形统计,根据分维数b₂的大小,将元素分为三类。Ⅰ类中数值小于2,包含了As、Co、Bi元素;Ⅱ类b数值范围在2～3之间,包含Sn、Ag、Zn、Pb、Cu、W元素,是主要成矿元素;Ⅲ类数值大于3,包含了Mn、Sb、Mo、Ti、V、Ni、Cr、Hg元素,并指出从分形曲线的拐点和间断性也可以判断矿区存在多期次成矿活动。3、在详细研究矿床的同生与后生特征的基础上,通过与国内外同类矿床的对比分析,并按成矿演化的时、空分布特征以及矿床的主导成矿作用,将老君山矿集区内矿床系统地划分为三大成矿类型:①喷流沉积—变质改造—岩浆热液叠加富集型成矿类型;②高温岩浆热液类型;③燕山晚期中低温热液型成矿类型。突破了传统的“唯花岗岩成锡”的观点,拓宽了找矿思路。4、从成矿系列研究出发,依据矿床成因及矿物组合特征,将老君山矿集区内所有矿化类型进行了重新分类,划分出七种与成矿类型相联系的矿床,即:层状锡石—硫化物矿床;层状锡多金属矿床;层状钨多金属矿床;长英岩脉型锡钨矿床;长英岩型锡钨矿床;石英脉型锡钨矿床、似层状铅锌矿床。并对各种类型矿床成矿地质条件、成矿作用、控矿规律以及地球化学特征等进行了详细的研究。5、将矿床成矿作用与壳体大地构造（递进）演化—运动相联系,详细阐述了老君山矿床成矿的多大地构造演化阶段、多成矿物质来源、多控矿因素、多成矿方式以及多种成矿作用的五“多”特点。建立了完整的矿床多因复成成矿（递进）演化综合模式,并分析了矿床多因复成成矿作用的内在机制。6、应用遥感信息提取技术对老君山地区进行遥感线性信息解译和矿化蚀变信息提取,进一步分析了老君山矿集区遥感信息成矿规律以及构造与蚀变的分形特征,并将其与其它地区线性构造对比,推测老君山矿集区断裂分形结构偏于复杂,活动性偏强,成矿规律更为复杂。7、利用空间分析方法和信息统计单元方法对老君山矿集区进行了数字矿床空间信息成矿预测模型建立的研究。研究得到了每个网格信息单元的成矿有利度,按成矿有利度0.5、0.6和0.7为异常分界点,对找矿预测单元进行了分级,预测单元可分为3级,即A级、B级和C级,其中A级预测单元（大于0.7）为成矿条件最有利,找矿标志明显,并具有寻找大型多金属矿床的潜力;B级预测单元（介于0.6和0.7之间）为成矿条件比较有利,找矿标志较明显,具有寻找中型多金属矿床的潜力;C级预测单元（介于0.5和0.6之间）为成矿条件较一般,但仍有找矿可能,具有寻找小型多金属矿床的潜力。8、为了检验矿床空间信息成矿预测模型的有效性,引入BP人工神经网络对成矿有利区进行评价。评价模型反演结果发现网络学习效果好,输出的值与期望的值满足评价要求,表明利用BP神经网络进行成矿有利区评价是可行的,从而也说明空间信息成矿预测模型得到的成矿有利度是正确的,成矿预测单元的划分是合理的。更多还原

【Abstract】 Laojunshan tin-zinc polymetallic ore-concentrating district, located in one of the mostest deposit-accumulated areas of Southeast Yunnan, is famous for its indium first reserves and one of the three tin producing base in China.Guided by the Diwa theories of Polygenetic and compound progressive metallogenic and metallogenic prognosis, starting with the geological geochemical features of the different type ore deposit （body） in Laojunshan deposit-accumulated area, reconstruction of protoliths of metamorphic rocks, metallogensis and extracting of remote sensing information etc, combined with the regional mineralization background,which summarized mineralization geological condition, trace elements distribution, characteristics of structural evolution, metallogenetic sequence and metallogenic evolution model. It has researched of the Ore-controlling mechanism, remote sensing interpretation, the geological modeling and spatial information metallogenic prognosis by adopting Polygenetic and compound progressive metallogenic theory, the mulfractal theory, the BP neural network evaluation model and spatial information metallogenic prognosis model etc. The main research findings and new viewpoints achieved in this paper are as follows:1. Through the ananlysis on the reveal features of the structure pattern , trackways and its suited structures, the author has fully and systematically explained the development and evolution features of the stress field and structure system in different stages, pointing out the importance of the Laojunshan vault,Wenshan-Malipo fracture and Maguan-Dulong fracture control over mineralization.2. By using multifractal method, 17 microelements have been studied on the base of factor analysis and factor analysis, according to the fractal dimension （b₂）,trace elements has been divided into three classes.Class I which are composed by As,Co,Bi. is less than 2. ClassⅡwhich are composed by Sn,Ag,Zn,Pb,Cu,W.is between 2 and 3. ClassⅢwhich are composed by Mn,Sb,Mo,Ti,V,Ni,Cr,Hg.is more than 3. Then the author has pointed out existing polyphase metallogenic movements from the estimation about inflection and discontinuity of the fractal curve in the ore field.3.On the basis of detailed study on syngenetic and subsequent characteristics of the mineral deposits and compareing with the similar type mineral deposits at home and abroad, The author has divided those mineral deposits in Laojunshan polymetallic ore-concentrating district into three deposit metallogenic types, according to the spatiotemporal distribution characteristics of the metallogenic evolution and dominant mineralization, i.e. Exhalative sedimentation-metamorphism alteration-magmatic hydrothermal superimposition and enrichment types, magmatic hydrothermal superimposition and alteration types of late Yanshan epoch, epithermal vein type in medium to temperature types of late Yanshan epoch, which breaking through the viewpoint of traditional "the monism of granite forming tin deposit", broadening the thought of ore prospecting.4. Starting with the study of deposit metallogenic series and according to the ore genesis and characteristic of mineral assemblage, the author has also reclassified all the deposits in Laojunshan ore-concentrating district into seven genetic types which associated with deposit metallogenic types, i.e. layered Cassiterite-sulfide deposit, layered Tin polymetallic ore deposit, layered tungsten polymetallic ore deposit, tin tungsten deposit of flesite vein, tin tungsten deposit of flesite, tin tungsten deposit of quartz vein and layered zinc-lead deposit. Furthermore, the author has detailedly studied the metallogenic conditions, mineralization, ore-controlling regularity and geochemical characteristics of all types of the deposits.5. Contacting the mineralization with the crustobody geotectonic evolution-movement features, the author has studied in detail the characteristics of multistage geotectonic evolution, multifactor mineralization control, multisource ore-forming materials, multicycle and multimode mineralization, set up a integrated Polygenetic compound mineralization evolution mode of the deposits, and studied its intrinsic mechanism of the mineral deposit Polygenetic compound mineralization.6.By using remote sensing information extraction method, remote sensing linear information and mineralization information have been studied. Further, remote sensing information ore-forming rule and fractal features of structures and alteration of Laojunshan deposit-accumulated area have been analyzed. After comparing with other ore fields, the author has guessed its fractal structure is inclined to complicated, activities is inclined to strong and ore-forming rule is more complicated.7.The space analysis method and the information statistics unit method have been used to research the model establishment of the spatial information metallogenic prognosis in Laojunshan deposit-accumulated area, which can compute the mineralization beneficial degree of each grid information unit, according to the mineralization beneficial degree, taking 0.7,0.6 and 0.5 as the dividing point of abnormity, the prospecting target has been divided the level of A,BandC. Level A units are relatively advantageous to the mineralization condition, are the obvious prospecting criteria and have the potential of seeking the large-scale polymetallic deposits; Level B units are the better prospecting criteria and have the potential of seeking the medium polymetallic deposits; Level C units are the poor prospecting criteria and have the potential of seeking small polymetallic deposits.8.In order to confirm the validity in the model of the spatial information metallogenic prognosis,the paper imports the BP artificial neural network to evaluate the mineralization beneficial area, The model inversion result discovers the studied effect of the neural network is good, the output value and the expectation value satisfies the evaluation requirement, which indicated it is feasible using the BP neural network to evaluate the mineralization beneficial area, thus also explained the computed mineralization beneficial degree is correct in the model of the spatial information metallogenic prognosis, the division of the prospecting units is reasonable.更多还原