摘要
简支转连续梁桥二次预应力钢束的合理配置对成桥后的应力储备和线形有着重要影响.通过有限元计算分析了二次预应力筋张拉参数(长度、根数、张拉控制应力)对结构的影响,改变张拉参数进行若干组有限元计算,生成BP神经网络样本,将其作为遗传算法寻优的子程序识别出最优的二次预应力筋张拉参数.襄樊四桥的工程实例亦表明,该方法获得了较为满意的优化结果.
The rational configuration of double prestressed steels of the continuous beam transferred from simply supported beam have important influence on stress storage and sag. Through the calculation of the FEM, the influence of double prestressed steels on the structure is analyzed according to the various parameters of the double prestressed steels; the sample of neural network will be obtained through the calculation of the FEM. Combined with the neural networks, genetic algorithm will identify the most excellent length of double prestressed steels. The engineering practice also shows that the method has acquired relatively satisfactory results.
出处
《武汉大学学报(工学版)》
CAS
CSCD
北大核心
2009年第1期128-131,共4页
Engineering Journal of Wuhan University
基金
湖北省交通厅科技项目资助(编号:鄂交科教[2004]570号)
关键词
简支梁
连续梁
二次预应力筋
神经网络
遗传算法
simply supported beam
continuous beam
double prestressed steel
neural networks
genetic algorithm