基于权重反分析的膨胀土分类云模型研究  

Study on Cloud Model for Expansive Soil Classification Based on Weight Back Analysis

作　　者：高岭[1] 李建朋[1] 母焕胜[1] GAO Ling;LI Jianpeng;MU Huansheng(Hebei Provincial Communications Planning and Design Institute,Shijiazhuang 050011,China)

机构地区：[1]河北省交通规划设计院,石家庄050011

出　　处：《路基工程》2018年第6期21-26,共6页Subgrade Engineering

基　　金：河北省交通运输厅科技项目:复杂地质条件下特长大断面隧道安全及灾变控制理论与方法(20130012)

摘　　要：针对云模型综合评价方法指标客观赋权问题,将权重反分析与云模型深度耦合,提出了基于权重反分析的云模型综合评价方法,并应用于膨胀土分类中。首先,在推导优化算法适应度函数的基础上,给出了权重反分析与云模型耦合的具体步骤,之后选用自由膨胀率、塑性指数和标准吸湿含水率作为评判指标,基于膨胀土类别评判实例,开展指标权重的反分析计算,从而实现了膨胀土分类云模型的客观赋权。最后,通过样本实例的回代评判和样本外土样的分类检验,证明了新建模型的可行性与有效性。研究表明,基于权重反分析的膨胀土分类云模型构建过程中主观性干扰因素较小,评价指标易于测试获取,且分类结果较为准确,为云模型综合评价方法的应用研究提供了新思路。In order to objectively determine the weight of indicators in the comprehensive evaluation method using cloud model,the weight back analysis and the cloud model were deeply coupled and applied to the classification of expansive soil.Firstly,the fitness function of optimization algorithm was derived,and then the specific procedures for coupling of weight back analysis and cloud model were given.Secondly,free expansion rate,plasticity index and standard absorption moisture were selected as the indicators for expansive soil classification,and then the weight of the three indicators was determined by back analysis calculation.Finally,this newly developed model was proved to be feasible and valid by the classification verification.The study shows that the proposed method for the coupling of weight back analysis and cloud model is characterized by easy access of evaluation indicators,minor interference of subjective factor and high classification accuracy,therefore,this method deserves further promotion and application.

关 键 词：指标权重 反分析 膨胀土 分类 云模型 

分 类 号：TU443[建筑科学—岩土工程]