基于改进量子遗传算法的模型交互修正方法  

Model interactive modification method based on improved quantum genetic algorithm

作　　者：向胜涛 王达 XIANG Sheng-tao;WANG Da(School of Civil Engineering,Changsha University of Science and Technology,Changsha 410114,China;School of Civil Engineering,Central South University of Forestry and Technology,Changsha 410004,China)

机构地区：[1]长沙理工大学土木工程学院,湖南长沙410114 [2]中南林业科技大学土木工程学院,湖南长沙410004

出　　处：《浙江大学学报（工学版）》2022年第1期100-110,共11页Journal of Zhejiang University：Engineering Science

基　　金：国家自然科学基金资助项目(51878072);湖南省研究生科研创新资助项目(CX20190661);湖南省科技创新计划资助项目(2020RC4049)。

摘　　要：针对传统有限元模型修正方法低效率、高成本且易陷入局部极值的缺点,提出基于改进量子遗传算法的静力多尺度有限元模型交互修正方法.依据量子计算理论,对量子比特矢量态进行实数编码,以改进量子旋转门实现旋转角自适应更新,引入量子全局干扰交叉、变异、灾变等遗传操作,设计改进量子遗传算法.以某钢-混组合梁桥为工程背景建立多尺度有限元模型,建立目标函数,对待修正区域进行分块处理.利用最大互信息系数对待修正参数进行筛选,给出目标函数权重,通过Python语言实现了基于改进量子遗传算法的静力多尺度有限元模型交互修正.结果表明,改进量子遗传算法相较于传统遗传算法、量子遗传算法具有更高的性能与精度,自动交互修正方法的效率较高,对材料弹性模量、厚度、车辆荷载等参数的修正与工程实际测试的情况基本吻合,目标函数修正结果相较于有限元计算的初始值,挠度误差降低至1.4%~14.3%,混凝土底板应力误差降低至2.6%~18.8%,钢梁应力误差降低至0%~11.1%.A static multi-scale finite element model interactive modification method based on improved quantum genetic algorithm was proposed aiming at the disadvantages of traditional finite element model modification methods, which are low efficiency, high cost and easy to fall into local extremums. The quantum bit vector states were encoded by real numbers according to the quantum computing theory, and the rotation angle was adaptatively updated by improving the quantum revolving gate. The improved quantum genetic algorithm was designed by introducing the quantum global interference crossover, mutation, catastrophe and other genetic operations. A multiscale finite element model was established with a steel-concrete composite girder bridge as the engineering background. The objective function was established, the correction region was partitioned, and the maximum mutual information coefficient was used to screen the parameters and obtain the weight of the objective function. The interactive modification of static multi-scale finite element model based on improved quantum genetic algorithm was realized by Python language. Results show that the improved quantum genetic algorithm has higher performance and accuracy than the traditional genetic algorithm and quantum genetic algorithm, and the automatic interactive modification method is more efficient. The modification of material elastic modulus, thickness, vehicle load and other parameters accorded with the actual engineering test. The deflection error was reduced to 1.4%-14.3%, the stress error of concrete floor was reduced to 2.6%-18.8%, and the stress error of steel beam was reduced to 0%-11.1% compared with the initial finite element calculation.

关 键 词：桥梁工程 钢-混组合梁 改进量子遗传算法 实桥试验 模型交互修正 

分 类 号：U441[建筑科学—桥梁与隧道工程]