基于不同指标无量纲化方法的岩爆预测模型优选

为研究不同指标无量纲化方法对岩爆等级预测模型精度的影响，提高岩爆预测准确率，选取应力系数、脆性系数和弹性能量指数作为预测指标。基于104组岩爆实例大样本数据，采用统一极差处理法、差异化极差处理法、平均化处理法和归一化处理法4种指标无量钢化方法，对预测指标的原始数据进行处理，建立不同的岩爆预测距离判别模型并进行工程实例应用。研究结果表明:基于平均化处理法的岩爆预测模型的回判准确率高达97.1%；对不同矿山、隧道和水电站的6个工程实例的预测结果符合实际情况，说明其是一种准确率高、方便实用的岩爆预测模型。

Optimization of rockburst prediction model based on different index dimensionless methods

In order to study the influence of different index dimensionless methods on the accuracy of prediction models on the rockburst grades,and improve the accuracy of rockburst prediction,the stress coefficient,brittleness coefficient and elastic energy index of rock were chosen as the prediction indexes.Based on 104 groups of large sample data of rockburst examples,the original data of prediction indexes were processed by using four types of index dimensionless methods including the unified extreme value processing method,differentiated extreme value processing method,averaging processing method and normalized processing method.Different distance discriminant models of rockburst prediction were established and applied in the engineering cases.The results showed that the accuracy of rockburst prediction model based on the averaging processing method reached as high as 97.1%,and the prediction results on six engineering cases of different mines,tunnels and hydroelectric power stations were completely consistent with the actual situation,which indicated that it is a rockburst prediction model with high accuracy,convenience and practicability.