集对理论聚类分析法综合预测煤与瓦斯突出

1.西安科技大学 能源学院,陕西 西安 710054; 2.西安石油大学 电子工程学院,西安 710065; 3.教育部 西部矿井开采及灾害防治重点实验室,陕西 西安710054

煤与瓦斯突出; 集对理论聚类分析法; 综合预测; 预测指标; 相似权计算法

Comprehensive forecast of coal and gas outburst on the basis of set pair theory and clustering analysis method
WEN Hu1,3,ZHAO Zhi-feng1,2,3,GUO Jun1,3

(1.College of Energy Science and Engineering,Xi'an University of Science and Technology,Xi'an 710054,China; 2.School of Electronic Engineering,Xi'an Shiyou University,Xi'an 710065,China; 3.Key Laboratory of Western Mine Exploration and Hazard Prevention,Ministry of Education,Xi'an 710054,China)

coal and gas outburst; set pair theory clustering analysis; comprehensive forecast; predicted index; similarity coefficient method

DOI: 10.13800/j.cnki.xakjdxxb.2015.0504

备注

基于集对理论和聚类分析相结合的方法,将现行的钻屑指标法、钻孔瓦斯涌出初速度法、电磁辐射法和单项指标法等单因素单数值预测煤与瓦斯突出的方法和指标,通过集对理论综合运用一个具体的集对联系度数学模型表达出来,结合系统聚类分析思想,通过计算预测系统联系度之间的同异反距离,按照同异反模式识别的“择近原则”,来辨别待测样本系统所属的类别,从而进行多因素综合预测煤与瓦斯突出。通过以焦作九里山矿巷道掘进工作面为例,应用综合预测模型计算分析煤与瓦斯突出预测指标的实例数据,结合相似权计算法确定各指标因素权重,提高预测模型精度,预测结果符合现场实际情况,取得了良好的预测效果。

Based on the set pair theory and the system clustering analysis method, combining the current numerical and single parameter prediction methods and indexes of coal and gas outburst, such as the drilling cutting index method, the borehole gas emission velocity method, and the electromagnetic radiation method. Through the set pair theory, integrated use of a specific connection degree mathematical model, identify the category of the sample system under test by calculating the difference and the distance between the contact degree of the forecast system, combined with the system clustering analysis, according to the difference and the pattern recognition of ‘choose the nearly principle', thus to multi-factor comprehensive prediction of coal and gas outburst. Put the mine roadway drivage face of the mount Jiuli in Jiaozuo as an example, it has obtained the good prediction effect and good effectiveness by the application of comprehensive prediction model to analysis and calculate the prediction index instance data of coal and gas outburst, combined with the similarity coefficient calculation method to determine the index weight, and improve precision of the prediction model, the prediction results conform to the actual situation.