基于STSNN聚类算法的用沙坝矿微震事件活动特征研究

发布时间：2018-03-07 23:07

本文选题：STSNN算法　切入点：聚类分析　出处：《中国安全生产科学技术》2017年02期 　论文类型：期刊论文

【摘要】：利用时空共享近邻聚类算法(STSNN)对用沙坝矿微震事件进行聚类分析,通过对噪声率进行有效性评价,最终确定k=6和ΔT=6为该算法的最佳输入参数,识别得到98个微震事件聚集区域,最大类簇有544个微震事件,并且该类簇主要集中在用沙坝矿的断层区域。对该类簇微震事件活动特征进行分析,主要包括微震事件的24 h分布、微震事件活动率、视体积、施密特数及劲度系数,根据活动规律的变化特征,提出微震活动率急剧下降并且累积视体积曲线忽然上升、施密特数和劲度系数先升后降的点作为岩体失稳发生破坏性事件的预警点。通过对微震事件活动规律的研究可为大事件的产生提供有效的预判信息,为保证矿山安全生产发挥重要的作用。
[Abstract]:The clustering analysis of microseismic events in Shaba mine is carried out by using space-time shared nearest neighbor clustering algorithm (STSNN). By evaluating the effectiveness of noise rate, the best input parameters of the algorithm are determined as kan6 and 螖 Tn6, and 98 microseismic event gathering areas are identified. There are 544 microseismic events in the largest cluster, and this kind of cluster is mainly concentrated in the fault area of the Shaba Mine. The characteristics of the microseismic events are analyzed, including the 24 h distribution of the microseismic events, the activity rate of the microseismic events and the apparent volume of the microseismic events. According to the changing characteristics of the law of activity, Schmitt number and stiffness coefficient, it is proposed that the rate of microseismic activity decreases sharply and the cumulative apparent volume curve rises suddenly. The points where the Schmitter number and stiffness coefficient rise first and then fall down are used as early warning points for the occurrence of destructive events in rock mass instability. Through the study of the activity law of microseismic events, effective prediction information can be provided for the occurrence of major events. In order to ensure the safety of mine production play an important role.
【作者单位】：中南大学资源与安全工程学院;
【基金】：中南大学中央高校基本科研业务费专项资金项目(2016zzts447)
【分类号】：TD311

【相似文献】