基于改进遗传算法的桥梁结构传感器优化布置  

OPTIMAL SENSOR PLACEMENT ON BRIDGE STRUCTURE BASED ON GENETIC ALGORITHM

作　　者：黄民水[1] 朱宏平[1] 李炜明[1] 

机构地区：[1]华中科技大学土木工程与力学学院

出　　处：《振动与冲击》2008年第3期82-86,共5页Journal of Vibration and Shock

基　　金：国家自然科学基金项目(50378041);教育部新世纪优秀人才支持计划(NCET-04-0720)

摘　　要：为了解决桥梁结构健康监测中的传感器优化布置问题,提出一种基于二重结构编码遗传算法的传感器优化布置方法。首先改进了编码方法,采用二重结构编码进行种群的初始化、交叉和变异,然后选择时采用最优保存策略,交叉时采用自适应部分匹配交叉,变异时采用自适应逆位变异。该法克服了传统遗传算法应用于大型结构时收敛速度慢且易陷入局部最优的缺陷,大大加快了收敛速度,并确保能够搜索到最优解。最后通过一个桥梁工程的实例分析,证明了该法在搜索能力、计算效率和可靠性方面明显优于序列法,可广泛地应用于桥梁结构的健康监测。An optimization approach based on coding genetic algorithms was developed for optimal sensor placement in health monitoring of bridge structure. The coding method was improved, in which the initialization, crossover and mutation are based on dual-structure coding, and the elitist strategy,adaptive partially matched crossover,and adaptive inversion mutation are adopted respectively in the proceses of selection,cross and mutation.The method overcomes the defects of slow convergence rate and easily getting stuck at a local optimum when traditional genetic algorithms are applied in health monitoring of bridge structure. Through a bridge example, the method was proved to have better search capability, computational efficiency and reliability than serial method.

关 键 词：传感器优化布置 遗传算法 桥梁结构 健康监测 二重结构编码 

分 类 号：U466.3[机械工程—车辆工程]