软土地基沉降预测的GM(2,1)模型分析和改进  

Analysis and Improvement of GM(2,1) Model for the Prediction of Soft Foundation Settlement

作　　者：牛欣欣[1] 

机构地区：[1]西安铁路职业技术学院土木工程系,陕西西安710600

出　　处：《安全与环境工程》2014年第2期139-142,共4页Safety and Environmental Engineering

摘　　要：目前大多数地基沉降预测模型都是统计模型,物理意义不够明确。本文以灰色系统预测模型中的GM(2,1)模型作为软土地基沉降预测模型,通过分析GM(2,1)模型与Asaoka公式之间的联系,指出其白化方程与Asaoka法二阶微分方程具有严格的一致性,为GM(2,1)模型在地基沉降预测中提供了明确的物理意义;利用最小二乘法和模式搜索法对GM(2,1)模型的参数求解进行了优化,并将GM(2,1)模型应用于实际工程中,结果表明该模型优于双曲线法和Asaoka法,可以应用于软土地基的沉降预测。Most existing prediction models in the forecast of foundation settlement are statistical models which do not have explicit meanings. This paper applies GM（2,1） model to forecasting foundation settlement. Firstly, the pa- per analyzes the relationship between GM（2,1） model and Asaoka method and points out that the second-order dif- ferential equation of grey model is equivalent to that of Asaoka method,which is the theoretical basis for GM（2,1） model in the forecast of foundation settlement. Secondly, the paper proposes an improved solution to GM（2,1） model by using least square method and pattern search method. Finally, the study applies GM（2,1） model to a case study. The results show that GM（2,1） model is superior to hyperbolic model and Asaoka method and verify the validity of GM（2,1） model in the forecast of soft foundation settlement.

关 键 词：软土地基 沉降预测 GM(2 1)模型 ASAOKA法 参数优化 

分 类 号：X948[环境科学与工程—安全科学] TU441.6[建筑科学—岩土工程]