基于Kriging模型的重力坝结构优化

Gravity Dam Structure Optimization Based on Kriging Model

为解决优化算法与有限元分析直接结合对坝体结构进行优化带来的计算量大、效率低的问题,以某混凝土重力坝非溢流坝段为分析对象,分别构建坝体安全系数、横断面最大宽度及横断面面积Kriging模型;结合遗传算法,以坝体横断面面积最小为优化目标对坝体进行优化。在此基础上,应用功能梯度材料思想对最优断面材料分区,使坝体关键部位应力分布更加合理。研究结果表明:(1)Kriging模型的计算精度满足工程设计要求,基于Kriging模型的优化算法较原算法大幅度缩短优化时间;(2)优化后坝体断面面积减小14.02%;(3)坝踵处采用弹性模量为14 GPa混凝土,坝趾处采用弹性模量为25.5 GPa混凝土且弹性模量梯度不大于11.5 GPa,有利于坝体应力分布,具有一定的工程应用价值。

In order to solve the problem of large computational complexity and low efficiency caused by the optimization of the optimization algorithm and the finite element analysis,the non-overflow dam section of a concrete gravity dam is taken as the analysis object,and the Kriging Model of safety factor and maximum width of cross-section and cross-sectional area of the dam are respectively constructed.Combined with genetic algorithm,the dam body is optimized with the minimum cross-sectional area of the dam as the optimization target.On this basis,the functional gradient material idea is applied to partition the optimal section material,which makes the stress distribution of key parts of the dam more reasonable.The results show that: ① The calculation accuracy of the Kriging Model satisfies the engineering design requirements,and the optimization algorithm based on the Kriging model greatly shortens the optimization time compared with the original algorithm; ② the area of the dam body after optimization has reduced by 14.02%; ③ the dam heel is made of 14 GPa concrete with the dam toe 25.5 GPa and the elastic modulus gradient is not more than 11.5 GPa which is beneficial to the stress distribution of the dam and has certain engineering application value.