基于神经网络的钢筋混凝土异型节点损伤分析

提出了将神经网络技术引入到钢筋混凝土构件损伤分析中的意义,并介绍了其在结构分析中的应用。根据低周反复荷载作用下钢筋混凝土框架异型节点的受力特点和破坏模式,定义了节点的损伤指标并分析了影响节点损伤指标的主要因素。利用前馈神经网络方法建立了该类节点的损伤模型。通过与传统分析方法相对比可知,在拥有了充分可靠的试验数据基础上将神经网络技术引入到伪静力条件下钢筋混凝土构件的损伤分析中可在一定程度上弥补传统计算方法的不足,为异型节点的研究提供了一种切实可行的途径。

Damage analysis of RC abnormal joints based on neural networks

The analysis of the damage to the reinforced concrete member by neural networks is conducted.According to the mechanical character and failure model of reinforced concrete abnormal joints under low reversed cyclic loading,a damage index of abnormal joints is put forward and a new damage model based on feed forward neural networks is presented.After comparing with the conventional methods,a conclusion is drawn,that is,in the condition of pseudo static,the neuro-computation based on sufficient test data can make up the deficiency of conventional methods and provides a useful means for the research of abnormal joints.