青海驼路沟矿区矿带构造特征及深部成矿预测研究  

Study on structural characteristics of ore belt and deep metallogenic prediction of Tuolugou mining area in Qinghai Province

作　　者：赵寿 韩生荣 李泽峰 Zhao Shou;Han Shengrong;Li Zefeng(No.3 Exploration Institute of Geology Resources of Qinghai Province,Xining 810029,China)

机构地区：[1]青海省第三地质勘查院勘查部,青海西宁810029

出　　处：《能源与环保》2025年第3期151-156,161,共7页CHINA ENERGY AND ENVIRONMENTAL PROTECTION

基　　金：2022年度青海省省级地质勘查专项资金项目(青自然资〔2022〕12号)。

摘　　要：针对青海驼路沟矿区深部成矿预测难题,采用青海驼路沟矿区矿带构造特征及深部成矿预测进行研究,以降低成本并提升找矿效率。首先分析了驼路沟矿区矿带的构造特征,揭示以韧性剪切与褶皱构造为主,矿带内强弱变形域交替出现,观察到“W”型和“N”型同斜褶皱及石英脉褶等构造现象。微观分析显示,韧性剪切带内发育糜棱岩结构,包含不对称长石残斑与新颗粒等特征性显微组构。采用粒子群算法优化支持向量机(SVM)模型参数,将优化后的SVM模型应用于矿带构造特征数据样本,构建深部成矿预测模型。通过输入详细的矿床地质信息,模型实现了对驼路沟矿区深部成矿潜力的预测。研究结果显示,模型精准预测了深部成矿预测节点的分布,共识别出43个有矿节点和107个无矿节点,提高了预测的准确性和效率。表明该方法为矿产资源的勘探与开发提供了明确的地质依据,不仅验证了基于粒子群算法优化的SVM模型在深部成矿预测中的有效性,还为驼路沟矿区地质环境的矿产勘查提供了新方法和实践指导,具有一定的科学意义和应用价值。In response to the difficult problem of deep metallogenic prediction in the Tuolugou mining area of Qinghai Province,the structural characteristics of the ore belt and deep metallogenic prediction in the Tuolugou mining area of Qinghai Province were studied to reduce costs and improve exploration efficiency.Firstly,the structural characteristics of the ore belt of Tuolugou mining area were analyzed,revealing that ductile shear and fold structures are the main ones,and strong and weak deformation domains alternate within the ore belt.Structural phenomena such as"W"-shaped and"N"-shaped synclinic folds and quartz vein folds were observed.Microscopic analysis shows the development of mylonite structures within the ductile shear zone,including characteristic microstructures such as asymmetric feldspar remnants and new particles.Using particle swarm optimization algorithm to optimize the parameters of support vector machine(SVM)model,the optimized SVM model was applied to the structural characteristics data samples of the ore belt to construct a deep metallogenic prediction model.By inputting detailed geological information of the ore deposit,the model achieved the prediction of deep metallogenic potential in the Tuolugou mining area.The research results show that the model accurately predicted the distribution of deep metallogenic prediction nodes,identifying a total of 43 mining nodes and 107 non mining nodes,improving the accuracy and efficiency of prediction.This indicates that the method provides a clear geological basis for the exploration and development of mineral resources.It not only verified the effectiveness of the SVM model optimized by particle swarm algorithm in deep metallogenic prediction,but also provided new methods and practical guidance for mineral exploration in the geological environment of the Tuolugou mining area,having certain scientific significance and application value.

关 键 词：青海驼路沟矿区 矿区矿带 构造特征 深部成矿预测 粒子群算法 支持向量机 

分 类 号：P618.62[天文地球—矿床学]