基于增量动力分析和人工神经网络计算结构影响系数和位移放大系数

Calculating R and C_d based on IDA and ANN

为考察不同的数据处理方式对计算结果的影响,利用ANSYS的瞬态动力分析功能对六层三跨的人字形、V形和SX形中心支撑钢框架进行增量动力分析(IDA).分别采用多项式最小二乘法和人工神经网络原理拟合增量动力分析的数据,得到结构的IDA能力曲线并计算结构影响系数和位移放大系数.结果表明:数据处理方式对结果有一定影响,利用人工神经网络原理进行结构IDA能力曲线的拟合,能够得到更真实的能力曲线.

In order to research the influence of different method of data handling on the computing results,the function of transient dynamic analysis in ANSYS program was used to carry out the Incremental Dynamic Analysis（IDA） of inverted V-Concentrically Braced Steel Frames（CBSFs）,V-CBSFs and SX-CBSFs with six layers and three spans.The method of least square of polynomial and the method based on Artificial Neural Networks（ANN） were used to fit the IDA data so as to get the IDA curves and calculating Response Modification Factor R and Displacement Amplification Factor Cd.The research indicated that the method of data handling would have some influence on the results and fitting the IDA curves based on ANN may help get more actual IDA curves.