多段交叉遗传算法在连续刚构桥测点优化中的应用  

Application of Multistage Crossing Genetic Algorithms by Optimal Sensor Placement in Dynamic Test of Continuous Rigid Frame Bridge

作　　者：刘来君[1] 倪富陶 孙维刚[1] 邵永军 李晓[1] 

机构地区：[1]长安大学公路学院,陕西西安710064 [2]陕西高速公路工程试验检测有限公司,陕西西安710086

出　　处：《重庆交通大学学报（自然科学版）》2016年第2期6-8,88,共4页Journal of Chongqing Jiaotong University(Natural Science)

基　　金：陕西省交通运输厅科研基金项目(13-25K)

摘　　要：阐述了基于动力检测的传感器优化布置准则和方法,提出了一种应用改进遗传算法,该算法改进了遗传算法交叉的规则,将原来的两点交叉改进为多段交叉。并将该方法具体应用到某大桥的动力检测中,通过计算对比分析,发现多段交叉的遗产算法有效的可以防止了传统遗传算法收敛过早、陷入局部最优解等现象,能更好的利用初始种群的多样性,多段交叉遗传算法计算结果要优于传统遗传算法。The improvement of genetic algorithm method based on optimal arrangement of sensor in dynamic test. was elaborated. The proposed improved method by using multistage crossing instead of two-point crossing improved the code of crossing in traditional genetic algorithm and was applied in the dynamic test of some continuous rigid frame bridge. Through calculation and comparative analysis,it is discovered that the application of multistage crossing method can effectively prevent the happenings of premature convergence and local optimal solution and better utilize the variety of initial population. Hence,the results by multistage crossing method are more superior to that of traditional generic algorithms.

关 键 词：桥梁工程 多段交叉 遗传算法 测点优化 

分 类 号：U446.3[建筑科学—桥梁与隧道工程]