基于太赫兹光谱技术的煤岩识别方法  

Coal-rock Identification Method Based on Terahertz Spectroscopy Technology

作　　者：王昕 赵端[2] 丁恩杰[2] 

机构地区：[1]江苏建筑职业技术学院,江苏徐州221008 [2]中国矿业大学物联网(感知矿山)研究中心,江苏徐州221008

出　　处：《煤矿开采》2018年第1期13-17,91,共6页Coal Mining Technology

基　　金：国家重点基础研究发展计划(973计划)资助项目(2014CB046300);江苏建筑节能与建造技术协同创新中心博士专项(SJXTBS1708);江苏省普通高校研究生科研创新计划资助项目(KYLX_1385)

摘　　要：煤岩识别是无人化开采中的重大问题,为了研究基于太赫兹谱的煤岩识别方法,首先应用太赫兹时域光谱技术对不同矿井采集来的煤岩样本进行了实验,得到了煤岩样本的太赫兹谱。然后计算出煤岩样本的折射率谱和吸收谱,并分别将其作为训练集和测试集。研究不同支持向量机参数对建模以及煤岩分类结果的影响。结果表明,通过合理的寻优方式,并利用煤岩的吸收特性和折射特性等样本的固有属性,可以建立稳定的THz-SVM的煤岩识别模型,从而可以快速稳定地区分出煤岩介质,这也意味着可以通过该方法来描述采煤机的截割状态(割煤/割岩),从而为采煤机自动调高提供依据。Coal-rock identification method is an important problem of minerless mining, in order to study terahertz spectroscopy tech-nology of coal-rock identification method, firstly, the coal-rock samples that collected from different mine was tested by terahertz time-domain spectroscopy technology, and then terahertz spectroscopy of coal-rock mass was obtained. And then refractive index spectrumand absorption spectrum were calculated, and then these were done as training set and test set, respectively. The influence that differ-ent support vector machine parameters to building model and coal-rock classification results were studied. The results showed that stabil-ity THz-SVM coal-rock identification model could be built according reasonable optimization method and using coal-rock absorptioncharacters and refraction characters and so on, so coal-rock media could be distinguished quickly and stability, so it meaning that cut-ting state （cutting coal / cutting rock） of coal cutter could be described, and provided the basis for coal cutter height-adjustment auto-matic.

关 键 词：太赫兹时域光谱 煤岩界面 煤岩识别 支持向量机 

分 类 号：TD67[矿业工程—矿山机电]