欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
在采用未确知聚类评价模型进行多指标分级评价时,常采用置信度识别准则作为待测对象的属性识别,该准则中置信度的取值由人为取定,当置信度取值不同时,得到的分级判定结果往往出现差异,甚至产生完全不同的判定结果。通过距离判别的思想将未确知聚类理论中的置信度识别准则进行改进,并运用到岩爆烈度的分级预测中。根据岩爆发生的主要影响因素,选取岩石单轴抗压强度σc、单轴抗拉强度σt、最大切应力σθ及岩石的弹性变形能指标Wet为岩爆主要影响因子。并以σc/σ1、σθ/σc、W（ey）为岩爆烈度等级评价因子,建立未确知测度模型,以距离判别改进后的属性识别方法进行分级预测,并与原置信度识别准则得到的判别结果进行分析和比较。为验证改进模型的实用性,以贵州开磷集团马路坪矿区为例,采用改进的未确知聚类模型对其岩爆烈度等级进行预测分析。结果表明,预测结果与实际情况基本吻合,证明采用改进后的未确知测度模型的判别结果不仅消除了由于置信度取值不同造成的判别结果误差,降低了人为主观因素的影响,而且具有较高的判别准确性和可行性。
英文摘要：
The paper is inclined to make an approach to the predictive liability of the rockburst classification based on the improved unascertained clustering model when using the unascertained clustering estimation model for multi-indicators evaluation,whereas the confidence recognition criteria are always to be used as an attribute recognition criteria.But in the said criteria,it is necessary to do the artificial choice of the values of confidence,though such a kind of artificial choice may lead to the different classification results,even the totally opposite results.To reduce such errors likely to be caused by the different choices of the confidence values as much as possible,we have managed to optimize the confidence criteria of the unascertained clustering theory on the basis of the distance discrimination methods in this paper.And then,we have adopted the optimized criteria to classify the rockburst intensity in the following steps.First of all,we have chosen the uniaxial compressive strength σc,uniaxial tensile strength σt,in-situ shear stress σθ and the elastic energy index of rock Wet in accordance with the characteristics of the rockburst and its main causes as the main influential factors.And,next,we have chosen three evaluation factors including σc/σt,σθ/σc and Wet to establish a model for the unascertained clustering prediction.Thus,in so-doing,we have obtained the unascertained measure functions of the rockburst classification and its figures based on the eighteen groups of engineering sample data.And,the third,we have worked out the weights of all the evaluation factors by the theory of entropy and the multi-indicator measurement in correspondence with the single index measure and weights.And,finally,we have predicted the classification of the rockurst intensity both by the optimized attribute recognition method and the confidence criteria with the different λ-values at the same time.As compared with the results of the two methods,the result of our prediction shows that the errors brought about by
同期刊论文项目
同项目期刊论文
期刊信息
位置：