钦杭成矿带浙江段典型区矿产资源（金矿）综合评价研究

发布时间：2018-10-15 09:00

【摘要】：近年来,随着全球工业化、城市化的飞速发展,人类对矿产资源的需求空前急迫,矿产资源的严重不足已经成为制约全球经济发展的巨大瓶颈。并且随着传统找矿工作的难度和成本增加,对矿产勘查技术提出了更高的要求。因此,利用计算机、GIS、图像处理等高新科学技术,探索一套定量化的矿产资源综合评价体系,对指导新矿隐伏矿的找矿方向、缓解全球的矿产资源危机、推动社会经济的发展具有重大的现实意义。本研究以钦杭成矿带浙江段的金矿典型区为研究区,基于对研究区成矿条件和成矿机制的充分认识,对收集和整理的地质、物探、化探、地形及遥感等多元地学数据进行GIS平台下的成矿信息挖掘,并应用证据权法和支持向量机法两种矿产评价技术对研究区的金矿成矿潜力进行综合评价。涉及的研究内容和成果认识主要包括：(1)通过对研究区成矿地质背景、区域控矿构造及剪切带型金矿特征等的充分认识和总结,对收集的地质图、地球物理图、化学异常图、DEM影像、遥感影像等多元地学数据进行GIS平台下的空间配准和格式统一,并深入挖掘和提取多元成矿有利因素。(2)鉴于研究区植被覆盖程度高的特点,摒弃了传统的着眼于蚀变弱信息提取的遥感找矿技术,创新地提出“遥感环境单元”的概念,认为遥感多波段信息反映了某一特定环境单元下地质、岩性、土壤、气候等综合因素的影响,而矿产的富集必然导致特有的遥感环境特征。因此,利用决策树分类技术,以分离出成矿有利的环境单元为目标,提取了遥感环境因子。(3)基于研究区矿产分布特征,提出“Wofe-SVM矿产综合评价方法”。根据WofE模型得出的权值系数及本研究提出的“成矿有利度综合评价指数”,对有利的矿产评价因子进行甄别和筛选。在获得优选的评价因子的基础上,利用SVM模型进行研究区的金矿综合评价,并利用WofE评价模型进行交叉印证和分析。最后结合多元成矿因素,对综合评价结果的准确性和可靠性进行分析。实践证明,研究区地勘资料丰富,多元成矿信息齐备,传统的基于专家经验的人工解译方式无法充分利用大量的地学信息,必须采用基于GIS的“知识驱动型”矿产资源综合评价技术进行信息挖掘和信息融合。并且当单一的评价方式不足以满足研究需要时,综合利用两种不同评价模型的优势,从两个维度对评价结果相互印证,保证了结果的准确性和可靠性。同时,充分利用遥感数据多波段信息,构建“遥感环境因子”,也是对传统地矿勘查工作的重要补充,对外围及深部找矿具有深远的意义。
[Abstract]:In recent years, with the rapid development of global industrialization and urbanization, the human demand for mineral resources is more urgent than ever. The serious shortage of mineral resources has become a huge bottleneck restricting the development of global economy. With the increasing difficulty and cost of traditional prospecting, higher requirements are put forward for mineral exploration technology. Therefore, using high and new science and technology such as computer, GIS, image processing and so on, the paper explores a set of quantitative comprehensive evaluation system of mineral resources, which can guide the direction of prospecting for hidden ore deposits of new mines and alleviate the global crisis of mineral resources. It is of great practical significance to promote the development of social economy. This study takes the typical gold deposit area of the Qinghang metallogenic belt as the study area. Based on the full understanding of the metallogenic conditions and metallogenic mechanism of the study area, the geological, geophysical and geochemical exploration of the collection and collation is carried out. The multivariate geological data, such as terrain and remote sensing, are used to mine metallogenic information on GIS platform, and two mineral evaluation techniques, evidence weight method and support vector machine method, are used to comprehensively evaluate the metallogenic potential of gold deposits in the study area. The contents and achievements of the research are as follows: (1) the geological maps and geophysical maps are collected by fully understanding and summarizing the metallogenic geological background, the regional ore-controlling structure and the characteristics of the shear zone type gold deposits in the study area. Chemical anomaly map, DEM image, remote sensing image and other multivariate geoscientific data are used to unify the spatial registration and format under the GIS platform, and further excavate and extract the favorable factors of multiple mineralization. (2) in view of the characteristics of high vegetation coverage in the study area, The traditional remote sensing prospecting technology focusing on the extraction of altered weak information is abandoned, and the concept of "remote sensing environmental unit" is put forward innovatively. It is considered that the multi-band remote sensing information reflects the geology, lithology and soil of a particular environmental unit. Climate and other comprehensive factors, and mineral enrichment will inevitably lead to unique remote sensing environmental characteristics. Therefore, remote sensing environmental factors are extracted by using decision tree classification technology to separate the favorable ore-forming environmental units. (3) based on the distribution characteristics of mineral resources in the study area, a comprehensive evaluation method of Wofe-SVM mineral resources is proposed. According to the weight coefficient obtained by WofE model and the comprehensive evaluation index of metallogenic advantage, the favorable mineral evaluation factors were screened and screened. On the basis of obtaining the evaluation factors of excellent selection, the comprehensive evaluation of gold deposits in the study area is carried out by using SVM model, and the cross-validation and analysis are carried out by using the WofE evaluation model. Finally, the accuracy and reliability of the comprehensive evaluation results are analyzed based on multivariate metallogenic factors. It has been proved by practice that the geological prospecting data in the study area are abundant and the multivariate ore-forming information is complete, and the traditional manual interpretation method based on the expert experience can not make full use of a large amount of geoscience information. The knowledge driven comprehensive evaluation technology of mineral resources based on GIS must be adopted for information mining and information fusion. And when a single evaluation method is not enough to meet the research needs, the advantages of two different evaluation models are comprehensively utilized to verify the evaluation results from the two dimensions, which ensures the accuracy and reliability of the results. At the same time, making full use of the multi-band information of remote sensing data to construct the "remote sensing environment factor" is an important supplement to the traditional geological exploration work, and has far-reaching significance to the peripheral and deep prospecting.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P618.51

【参考文献】