论文部分内容阅读
距离判别法是一种按多指标性能参数进行分类的方法,已被广泛应用在岩爆等级分类中去,但传统的距离判别分析方法对各个指标的重要性是同等对待的,夸大了一些微小变化指标的作用.采用加权马氏距离对样本中各指标进行加权处理以区分其重要性.首先对岩爆主控因子进行主成分分析,计算出一组新的指标参数和与其对应的反应指标信息的贡献率,以贡献率组成对角矩阵作为权值矩阵,然后把数据进行标准化处理,计算加权马氏距离判别岩爆等级,并与马氏距离判别法进行了比较.比较结果表明:加权距离判别法具有更高的判别能力,判别效果更符合实际情况.利用建立好的加权马氏距离模型,对马路坪硬岩矿山和铜陵冬瓜山深埋硬岩矿山岩爆进行预测.实例研究表明:预测结果与工程实例实际岩爆情况相符合.
The distance discrimination method is a method of classifying according to multi-index performance parameters and has been widely used in rockburst classification. However, the traditional distance discriminant analysis method treats the importance of each index as equal, exaggerating some minor We use the weighted Mahalanobis distance to weight each index in the sample to distinguish its importance.Firstly, the principal component analysis of the main controlling factor of rock burst is carried out to calculate a new set of index parameters and their corresponding reaction indexes The contribution rate of information, the contribution rate composed of diagonal matrix as the weight matrix, and then the data were normalized to calculate the weighted Mahalanobis distance to determine the level of rockburst and Mahalanobis distance discrimination method were compared.The results showed that: Discriminant method is more discriminative, and the discriminant effect is more in line with the actual situation.Based on the established weighted Mahalanobis distance model, the rock burst of Maluping hard rock mine and Dongguashan deep hard rock mine in Tongling is predicted.Examples show that : The prediction result is in accordance with the actual rock burst situation of the engineering example.