岩爆预测的多维云模型综合评判方法

A multidimensional cloud model for rockburst prediction

岩爆是地下空间开发和矿业工程中遭遇的重大挑战,其倾向性预测是一个迫切需要解决的难题。针对其预测过程中的不确定性,在一维云模型理论的基础上提出了基于多维云模型的岩爆倾向性预测方法。将各个预测指标分别作为多维云模型的一维变量,根据云理论和岩爆预测标准,选取合适的云模型数字特征,考虑指标的信息量和相关性,改进CRITIC法确定预测指标权重,生成综合所有预测指标的对应于各个岩爆等级的多维云模型,输入指标实测值得到岩体隶属于各个岩爆等级的综合确定度,依据最大确定度原则判定岩体岩爆等级。以国内外20组典型岩爆实例数据验证本文模型的合理性和有效性,并与一维云模型、RS-TOPSIS方法的预测结果进行对比。结果表明,多维云模型应用于岩爆倾向性预测是有效的,其可以直观、快速判定岩体的岩爆等级,为岩爆倾向性预测提供了一种新方法。

Rockburst is a significant challenge to the development of underground space and mining engineering. The prediction of rockburst is an urgent issue need be solved. Based on the one-dimensional cloud theory,a multidimensional cloud model for prediction of rockburst is proposed to deal with the uncertainty in the forecasting process. It regards each prediction index as a one-dimensional variable in a multidimensional cloud model,chooses the reasonable numerical characteristics of each cloud model according to the cloud theory and the classification standard for prediction of rockburst,and determines the weighting coefficient for each prediction index with the improved CRITIC method in view of the information and correlation of the indexes. Thus,the multidimensional cloud models belonging to each rockburst level that synthesizes all prediction indexes are generated and the final rockburst level of rock mass is determined according to the comprehensive certainty with the values of each index input into the models. Then the model proposed in the present paper is validated with the data of 20 groups of typical rock engineering at home and abroad. The predicted results of the proposed model are found to be consistent with the actual rock status and are compared with those from the related method,one-dimensional cloud model and the technique for order preference by similarity to ideal solution combined with rough set（RS-TOPSIS）. The results show that the multidimensional cloud model is effective and determines the rockburst level quickly.