文章收集郑西高速铁路地基黄土典型湿陷性试验资料(包括现场大型浸水试验及室内试验),以影响黄土湿陷系数主要因素为基础,运用MATLAB建立黄土湿陷系数的自适应神经网络模糊推理系统(ANFIS)预测模型。通过对样本的训练和预测,表明该模型预测结果与实际黄土湿陷系数十分接近。用多元线性回归法对这些非母体样品进行预测检验,经过对比ANFIS法优于多元线性回归法,证明ANFIS法是一种比较理想的预测方法。
Based on the main factor influencing the loess collapsibility, the data of collapsibility test on collapsible loess foundation in Zheng-Xi High-Speed Railway are collected and summarized, a model of ANFIS based on the index about the loess collapsibility has been established in this paper. Compared the results with the conclusion by the method of multivariate linear regression, it is shown that the ANFIS prediction model could improve the accuracy of forecasting. The results shows that the prediction result after trained and predicted for the samples is in good agreement with the experimental results, and this method is a relatively optimum method for forecasting, at the same time it provides an important viewpoint to explore the collapsibility law on the loess foundation of high-speed railway.
[1] 李景宜.陕西渭河下游湿地环境风险因素分析[J].地理科学,2007,27(3):371~375.
[2] 王家鼎,惠泱河.黄土地区灌溉水诱发黄土滑坡群的研究[J].地理科学,2002,22(3):305~310.
[3] 王家鼎,谢婉丽,骆凤涛.高填方加筋黄土路堤稳定性的有限元分析[J].地理科学,2007,27(2):268~272.
[4] 王家鼎,张倬元,李保雄.黄土湿陷的脉动液化机理[J].地理科学,1999,19(3):271~276.
[5] 冯连昌,郑晏武.中国湿陷性黄土[M].北京:中国铁道出版社,1980:141~149.
[6] 中华人民共和国建设部.中华人民共和国国家标准——《湿陷性黄土地区建筑规范》(GB50025-2004)[S].北京:中国建筑工业出版社,2004:12~19.
[7] 张利生.湿陷性黄土试验方法探讨[J].岩土力学,2001,22(1):207~210.
[8] 谢婉丽,王家鼎.模糊信息优化方法在黄土湿陷性评价中的应用[J].西北大学学报(自然科学版),2005,35(1):95~99.
[9] 刘 悦,王家鼎.黄土湿陷性评价中的模糊信息优化处理分析方法[J].西北大学学报(自然科学版),2000,30(1):78~82.
[10] 高国瑞.黄土湿陷变形的结构理论[J].岩土工程学报,1990,12(4):1~9.
[11] Miao T D. Deformation mechanism of collapsible loess in consideration of the microstucyure instability[J].Catena,1991,20(5):93-105.
[12] 李雨浓.影响黄土湿陷系数因素的分析[J].世界地质,2007,26(1):108~113.
[13] Takagi T,Sugeno M..Fuzzy identification of systems and its applications to nodeling and control[M]. IEEE Transon SMC,1985.
[14] 闻 新,周 露,李东江,等.MATLAB模糊逻辑工具箱的分析与应用[M].北京:科学出版社,2001.
