基于主成分物元可拓模型的岩爆倾向性预测

Prediction of rock burst proneness based on principal component matter-element extension model

岩爆会对井下工作人员和设备造成严重的伤害和破坏,开展岩爆倾向性预测研究具有重要的现实意义。以岩爆倾向性预测为目标,综合考虑岩爆的影响因素,从岩性、应力、围岩3个角度选取12个影响因素作为评价指标,采用主成分分析法和物元可拓模型建立新的预测模型即主成分物元可拓模型,选取国内7个典型工程实例,运用主成分物元可拓模型进行岩爆倾向性等级预测,将预测结果与实际情况进行对比,发现预测结果与实际情况基本吻合,准确率达到了85.71%。以白鹤滩水电站地下厂房为工程实例,对该预测模型的精度作进一步检验,结果表明,该水电站地下厂房的岩爆倾向性等级为Ⅱ级,属于弱岩爆,预测结果与实际情况吻合。

As rock burst is bringing about serious injury and destruction to underground workers and facilities,it is of great practical significance to carry out the prediction of rock burst proneness.Aiming at predicting rock burst proneness,twelve factors in terms of lithology,stress and surrounding rock were selected as the evaluation criteria,and a novel prediction model namely principal component matter-element extension model was established via principal component analysis based on matter-element extension model.With selection of 7 domestic typical project cases,the principal component matter-element extension model was employed to predict the level of rock burst proneness,from comparison it was found that the results from prediction were basically consistent with the physical truth and the accuracy was 85.71%.Further verification on the accuracy of the prediction model was performed with the project reference of underground powerhouse of Baihetan Hydropower Station.The results show that the rock burst proneness of underground powerhouse of the hydropower station was grade Ⅱ,which belongs to weak rock burst.The prediction results accorded with the actual conditions.