节点文献
深水桩墩结构振动台试验及地震响应预测分析
Experimental investigation of shaking table on pile-pier structure in deep water and prediction analysis of seismic response
【摘要】 动力模型试验是研究桥梁结构抗震设计理论的重要方法,而神经网络技术对非线性系统具有很好的辨识和预测功能.为了分析地震动作用下动水压力对结构的影响及探索神经网络应用于地震响应预测分析的可能性,进行了水下桩墩结构振动台模型试验及其仿真预测,衡量了水下桩墩结构的地震响应和动力特性.首先,介绍了相似律的选取、模型制作、试验现象及试验结果分析;然后,基于神经网络的预测功能,对模型试件的地震响应进行预测,并与试验结果对比研究;最后,分析试验结果及预测误差.试验结果表明:结构周围水体的存在改变了结构的地震响应及动力特性;训练有素的神经网络模型可以作为一个有用的工具,用于结构的地震响应预测.
【Abstract】 The dynamic model test is an important method to research into seismic design theory of the bridge structure.The function of identification and prediction of neural network can be applied to nonlinear system effectively.In order to develop more advanced and reliable design procedures and explore the possibility of application of neural network to analysis and prediction of seismic response,investigation and prediction on shaking table model test of submerged pile-pier system,including pile-pier and the lumped mass,are conducted.Firstly,the similitude laws,model making,failure process and experimental results are introduced.Then,based on the function of prediction of neural network,the seismic response of model specimen is predicted and compared.Finally,the test results and prediction error are analyzed.The experimental results show that the dynamic characteristics and seismic response of the specimen can be changed because of the effect of water;the neural network can predict the seismic response accurately,and it can be used as an effective supplement for the experimental research.
【Key words】 neural network; prediction; submerged pile-pier system; experimental research;
- 【文献出处】 大连理工大学学报 ,Journal of Dalian University of Technology , 编辑部邮箱 ,2013年01期
- 【分类号】U442.55
- 【被引频次】5
- 【下载频次】176