摘要
由于影响复合地基承载力的因素众多,各因素又互相作用,很难建立准确的承载力计算公式,现在设计中广泛使用的公式及其他理论公式仍是借助于试验基础上的半经验半理论公式,在实际工程中设计往往偏于保守造成不必要的浪费。文中提出利用MATLAB工具箱建立的基于遗传神经网络的CFG桩复合地基承载力预测模型来对CFG桩复合地基承载力进行预测,从预测的结果可知该方法运行简单,操作方便,准确度高。
Due to the multi-influence interacted factors on bearing capacity of composite foundation,it is hard to set up accurate calculation formula of the foundation capacity.Now the formula and other theory formula widely used in design are the partly experimental,partly theoretical formula based on tests,which are conservative and wasteful in engineering design.Based on MATLAB genetic toolbox,the neural network based prediction model is established and applied to CFG pile composite foundation bearing capacity prediction.It is shown that this method is easy to work and convenient to operate,and has high accuracy.
出处
《世界地震工程》
CSCD
北大核心
2010年第S1期263-266,共4页
World Earthquake Engineering
基金
黑龙江省教育厅科学技术研究项目(11513015)
