煤层冲击倾向性变权重可变集分类识别

1.淮南职业技术学院 技工部,安徽 淮南 232001; 2.安徽淮南矿业技师学院,安徽 淮南 232035

冲击地压; 冲击倾向性; 可变集识别模型; 变权模型; 博弈论; 类别特征值

Integrated variable weight and variable sets classification recognitionmethod for rock burst tendency of coal seam
GUO Jin-dong1,2,ZHU Yun-hui1,2

(1.Advanced Technical School,Huainan Vocational Technical College,Huainan 232001,China; 2.Huainan Mining Technician College,Huainan 232035,China)

rock burst; bursting liability; variable sets recognition model; variable weight model; game theory; grade feature value

DOI: 10.13800/j.cnki.xakjdxxb.2017.0316

备注

针对煤层冲击倾向性评价的模糊不确定性问题,为准确合理地评判冲击倾向性强弱,提出变权重可变集分类识别模型。模型选取动态破坏时间、弹性能量指数、冲击能量指数和单轴抗压强度作为分类指标,采用CRITIC法计算指标客观常权,利用指标数据与指标分类标准值区间的关系求解指标客观变权,最后运用博弈论对主客观权重进行优化集成,得到评价对象的变权权向量。引入可变集理论合理确定评价指标对各类别指标标准值区间的相对差异度,建立冲击倾向性可变集评价模型,依据类别特征值判断评价对象的隶属类别。实例表明,该方法较好地考虑了模型参数与样本指标权重的可变性,能科学合理地评价煤层冲击倾向性,识别结果更为可信,为煤岩冲击倾向鉴定与冲击地压危险性预测提供了新的方法和思路。

Aiming at the fuzzy uncertainty problems for rock burst tendency evaluation of coal seam,a classification recognition method using variable weight and variable sets was proposed to improve the recognition accuracy of bursting liability.Four indexes were selected as classifying indexes including duration of dynamic fracture,elastic strain energy index,bursting energy index and uniaxial compressive strength.The fixed weight of evaluation index was gained based on CRITIC,and variable weight of sample index was calculated by index data and standard value of index classification.Then the game theory optimization model was introduced to determine variable weight vector of evaluation sample considering subjective weights and objective weights.The relative difference degree between each index and standard interval of each level index was reasonably calculated by variable sets theory.Variable sets model was established to identify sample grade by stable grade feature value.The results show that the proposed approach can be applied to evaluate the rock burst tendency accurately based on the change of model parameters and index variable weight,and the assessment result can be more reliable.It provides a new theory and approach for identification of bursting liability and prediction of rock burst.