深厚冲积层人工冻土蠕变参数的模糊遗传反演  

Fuzzy genetic inversion of creep parameters of artificial frozen soil in deep alluvium

作　　者：姚亚锋[1,2] 荣传新[2] 程桦[2,3] 徐广舒[1] 

机构地区：[1]南通职业大学建筑工程学院,江苏南通226001 [2]安徽理工大学安全科学与工程博士后科研流动站,安徽淮南232001 [3]安徽大学资源与环境工程学院,安徽合肥230022

出　　处：《冰川冻土》2017年第3期634-640,共7页Journal of Glaciology and Geocryology

基　　金：国家自然科学基金项目(51374010;51474004);南通职业大学自然科学研究项目(16ZK05);江苏省高等学校大学生创新创业训练计划项目资助

摘　　要：冻土的蠕变特性对深井冻结法施工至关重要。针对某矿区人工冻土在-5℃、-10℃、-15℃和-20℃下进行单轴抗压强度试验,发现冻土的抗压强度受冻结温度变化影响,两者间为线性反比例关系。通过小生境原理对传统的遗传算法作模糊随机改进,给出模糊遗传算法的步骤思路,进而运用该算法反演冻土蠕变模型中的参数值,获得各温度下的蠕变模型。试验结果表明:蠕变模型计算值在蠕变各阶段与试验值吻合较好,准确反映了冻土蠕变的整体规律。可见,模糊遗传算法能有效反演蠕变参数,较传统方法更符合工程实际。Creep characteristics of frozen soil are the key to freezing method construction of deep shaft.In the mining area,uniaxial compressive strength tests are conducted on the artificial frozen soil at-5 ℃,-10 ℃,-15 ℃ and-20 ℃.It is found that the compressive strength of artificial frozen soil has a linear relation with temperature,on the whole,when the temperature is decreasing,the uniaxial compressive strength will be increasing.On this basis,traditional genetic algorithm has been fuzzy random improved by niche theory and obtained a new thought,which was used to improve fuzzy genetic algorithm to inverse creep model parameters of artificial frozen soil and obtain corresponding creep model.Experiments show that creep model can accurately reflect the frozen soil creep law because the model calculation values fit the experiment values well in each creep stage.Thus,one can see that fuzzy genetic algorithm is able to inverse parameters effectively,which is more suitable for engineering practice as compared with the traditional inversion algorithm.

关 键 词：人工冻土 蠕变参数 模糊遗传算法 参数反演 

分 类 号：TU445[建筑科学—岩土工程]