Rock mass explosibility is an important index to measure the difficulty of rock mass blasting,and an accu-rate evaluation of rock mass explosibility can provide a basis for reasonable blasting design.In this paper,rock density,uni-axial compressive strength,rock tensile strength,rock brittleness index,dynamic load strength,and integrity coefficient are selected as the indicators of rock mass explosibility data set.The data set of rock mass explosibility is standardized by Z-Score,and the influence of dimension on model prediction is eliminated.Naive Bayes,support vector machine,and XGBoost models are used to classify rock mass explosibility.The results show that XGBoost model can accurately evaluate rock mass explosibility and provide a new method for rock mass explosibility evaluation.
关键词
爆破/岩体可爆性/可爆性分级/XGBoost/机器学习算法
Key words
blasting/rock mass explosibility/explosibility classification/XGBoost/machine learning algorithm
维普
万方数据