自适应神经模糊系统的地基承载力组合模型

Composite model of foundation bearing capacity based on adaptive neural-fuzzy inference system

碎石桩为处理软基的常用方法之一,其承载力除用静荷载试验确定外,一般用经验公式计算。但目前各公式采用的方法与理论依据不同,涉及的物理力学参数各有所侧重,难以反映诸多因素对承载力的影响。为克服单个模型(经验公式)的局限性与不确定性,在MATLAB环境下,采用非线性映射能力的自适应神经模糊推理系统(ANFIS)建立了碎石桩复合地基承载力组合模型。实例分析表明,组合模型预测相对误差小于3%,优于各子模型,其精度满足工程应用要求,为确定地基承载力提供了一种新的方法。

The gravel pile is one of the methods of tackling soft-foundation. Besides by static load test, its bear-ing capacity is often calculated by empirical formulae. Because of the difference of methods and theories of theseformulae, and the bias of concerning physicomechanical parameters, the effect on bearing capacity of many fac-tors is difficult to be reflected, thus it affects the computation accuracy. In order to overcome the limitation andindeterminateness of single model (empirical formula), a composite model of foundation bearing capacity is givenbased on adaptive nonlinear neural-fuzzy inference system (ANFIS) in MATLAB. The practice shows the pre-diction relative error of the composite model is not only less than 3 % being superior to all of the sub-models, butits accuracy may meet the need of practical projects, thus it provides a new method for calculating foundationbearing capacity.