基于InSAR技术和SA-SVR算法的矿区沉降预测模型

为了解决矿区沉降预测模型精度低、预测模型与实际开采情形不符的问题，提出了一种基于合成孔径雷达干涉测量（Synthetic Aperture Radar Interferometry，InSAR）技术、支持向量回归算法（Support Vector Regression，SVR）以及模拟退火算法（Simulated Annealing，SA）相结合的新型矿区沉降预测模型。首先，以InSAR技术获取矿区沉降监测数据，对数据进行处理得到测试点的累计沉降量，并将其与GPS实际测量结果进行比较，发现二者吻合性较好。然后，进行矿区沉降预测模型构建，通过SVR算法得到静态沉降预计模型，再利用SA算法得到模型中的参数最优取值。为了使预测数据符合矿区开采实际情况，引入嵌入维数公式，得到矿区沉降预测动态模型及精度评价指标。最后，将构建的沉降动态模型应用于陕西省大柳塔矿区，得到预测值和实际监测值之间绝对误差的最大值为9 mm，相对误差的最大值为3%；模型评价指标通过计算得到试验区平均绝对误差的最大值为2.5%，最小的相关性指数为0.8，表明该模型预测精度较高。

Research Progress on Microbial Leaching Technology of Non-ferrous Metal Tailings

Microbiological leaching technology is one of the most promising technology in the development and utilization of valuable metal minerals in the mass stack of non-ferrous metal mine tailings in China， the dissolution of toxic metal elements， the full utilization of resources and the reduction of pollution. In order to systematically demonstrate the progress of microbiological leaching technology， the mechanism of microbiological leaching was introduced from the aspects of direct leaching， indirect leaching， galvanic effect and extracellular polymer（EPS） action； The key technology affecting the leaching effect were summarized from the leaching strains and external factors（including temperature， pH， redox potential and nutrients）； The role of tailings pretreatment， mutation breeding， and use of catalysts in enhancing the leaching effect was introduced. Finally， the development direction and prospect of microbiological leaching technology in the harmless and resource-oriented treatment of non-ferrous metal mine tailings were forecasted.