基于SVM的煤岩破裂与失稳预测模型  

Forecast model based on SVM during coal crack and destabilization

作　　者：刘叶玲[1] 张海燕[1] 来兴平[1] 

机构地区：[1]西安科技大学,陕西西安710054

出　　处：《煤田地质与勘探》2007年第3期62-65,共4页Coal Geology & Exploration

基　　金：国家自然科学基金项目(10402033);教育部西部矿山开采与灾害控制重点实验室重点项目(04JS19)

摘　　要：煤岩破裂与失稳的定量化实验是准确预测预报采空区动力学灾害的基础研究。基于采空区上覆煤岩破裂与动力学失稳相似材料模拟实验,利用支持向量机方法对煤岩破裂的不同阶段进行预测,找出大尺度采空区煤岩介质断裂与突然失稳的应力-位移-声发射“预警值(区)”,对开采中诱发动力灾害的危险源辨识及预警提供了定量化的依据,为工程现场准确、及时预报采空区煤岩断裂失稳规律提供了有效依据。Coal crack and destabilization quantitative experiment is a fundamental study on forecast dynamic hazard of gob area. Based on similarity material simulation experiment during coal crack and dynamic destabilization of gob area, forecasted different phases during coal crack by SVM technique, found out alarm value for stress-displacement-acoustic emission during coal medium crack and sudden destabilization of large scale gob area. By applied quantitative criteria for danger source discrimination and alarm induced dynamic hazard during mining, predicted rule timely and accurately for coal crack and destabilization of gob area. It provided for efficient guidance of the mining.

关 键 词：支持向量机 支持向量回归机 核函数 相似模拟实验 

分 类 号：TD315.3[矿业工程—矿井建设]