岩体可爆性等级判别的随机森林模型及R实现

Random Forest Model of Rock Mass Blastability Grading and R Language Implementation

在引入随机森林分类方法的基础上,选取岩石的容重、抗拉强度、动载强度及岩体完整性系数等指标,建立岩体可爆性等级判别的随机森林模型;在R语言环境下,编写模型数据处理、计算与结果输出的R语言代码,实现可爆性等级判别随机森林模型的计算,得出岩体可爆性等级判定的混淆矩阵,分析各指标对岩体可爆性分级的重要性。研究结果表明:岩体可爆性分级为三、四、六级的判别准确率可达100%;岩石容重对岩体可爆性等级判别的影响最大。这一结果证明了岩体可爆性等级判别的随机森林模型是可靠的,且具有较高的准确率。

On the basis of random forest classification method,some indexs of rock mass blasting grading such as density,tensile strength,dynamic load strength and other rock mass integrity coefficient are used in order to establish the random forest model of the rock mass blasting grading. In R language locale,the model data processing,calculation and outputs the result of R language code are written to achieve the calculation of random forest model of rock mass blastability grading and the confusion matrix of rock mass blastability rating are obtained. At the same time,the importance of each indicator for rock mass blastability classification is analyzed. The results show that the prediction accuracy rate of rock mass blasting classification of three,four,six prediction is up to 100%; The rock density is the greatest factor in rock mass blasting grading. The results of rock mass blasting grading prove that random forest model of rock mass blastability grading is reliable and has high accuracy.