基于均匀设计与神经网络的桥梁设计参数优化

摘要

In order to establish the mapping relationship between the bridge design parameters and the bridge performances and optimize the bridge design parameters,a method based on uniform design and neural network is presented.The strength,deflection and concrete dosage of bridge as taken as the optimization targets,and such main design parameters as the ratio of side to main span,middle beam height,pier top beam height and curve power are optimized.Sample data for optimization design are obtained by means of uniform experimental design,and BP neural network is employed to train sample data and establish the mapping relationship between the parameters of the bridge and the bridge performances,then the best design parameters for the bridge are calculated.%为了建立桥梁设计参数与桥梁性能之间的映射关系并对桥梁设计参数进行优化,提出一种基于均匀设计与神经网络的方法;以桥梁跨中截面的挠度、全桥最大应力、混凝土用量作为优化目标;边中跨比、跨中梁高、墩顶梁高和梁底曲线次数作为主要结构设计参数进行优化.通过均匀试验设计方法得出优化设计的样本数据,通过BP神经网络训练样本数据,建立桥梁参数与桥梁性能之间的映射关系,并基于该映射计算出桥梁最优性能状态下对应的设计参数,最后完成参数优化.