辽宁工程技术大学学报（自然科学版）2017，Vol.36Issue(4)：366-370,5.DOI:10.11956/j.issn.1008-0562.2017.04.006
基于主成分灰关联的瓦斯涌出量预测模型
Forecast of gas emission amount based on principle component and gay correlation
摘要
Abstract
To determine the nonlinear characteristics of gas emission,complex correlation among indexes and accurately forecast gas emission in a coal mine,based on principle component analysis and grey relational theory,the main indexes were obtained without collinearity and correlation;by any nonlinear function in the whole region,the support vector machine model coupled nonlinear combined model were built to forecast gas emission based on the distance model and the gay model,and the parameters were tuned by training samples set and the principle of absolute percentage error minimization,the absolute gas emission in a certain mining workface of a certain coal mine in Shenyang was analyzed.The results show that the collinearity which is between individual in production and extraction rate and other indexes is relatively strong,the collinearity which is between individual in coal seam pitch,closely the layer thickness and layer lithology and other indexes is relatively poor;the proposed method had lower errors than other original individual forecast method,the max absolute gas emission is 5.83％.关键词
瓦斯涌出量/非线性组合预测/主成分分析/灰关联/支持向量机Key words
gas emission amount/nonlinear combination forecast/principle component/grey system/support vector machine分类
资源环境引用本文复制引用
权国林,赵琳琳,邵良杉,温廷新,章菲菲..基于主成分灰关联的瓦斯涌出量预测模型[J].辽宁工程技术大学学报（自然科学版）,2017,36(4):366-370,5.基金项目
国家自然科学基金项目(71371091) （71371091）
辽宁省高等学校杰出青年学者成长计划项目(LJQ2012027) （LJQ2012027）
