基于多目标遗传算法的传感器优化布点研究-工程力学.pdf

第24 卷第4 期 Vol.24 No.4 工 程 力 学 2007 年 4 月 Apr. 2007 ENGINEERING MECHANICS 168 文章编号：1000-4750(2007)04-0168-05 基于多目标遗传算法的传感器优化布点研究 * 1 1 2 张连振 ，黄 侨 ，王潮海 (1. 哈尔滨工业大学交通科学与工程学院，黑龙江，哈尔滨 150090；2. 吉林省高速公路建设局，吉林，长春 130021) 摘 要：针对目前桥梁模态试验中传感器布点优化都是基于单一准则下的研究的局限性，提出采用双准则或多准 则下的传感器优化布点研究，该方法改进了以往布点优化中单一目标最优的限制，转而寻求多准则下的优化布点 的满意解，而非真正严格数学意义上的最优解。采用Pareto 遗传算法，设计了相应的遗传算子和编码方案，成功 求解了双准则下传感器最优布点问题，优化结束时，成功给出了一组pareto 最优解的前沿面，进而证明了提出的 方法是可行的。 关键词：模态识别；传感器；优化布点；遗传算法；多目标优化 + 中图分类号：U448.21 7; O34 文献标识码：A OPTIMAL SENSOR PLACEMENT BASED ON MULTI-OBJECT GENETIC ALGORITHM * 1 1 2 ZHANG Lian-zhen , HUANG Qiao , WANG Chao-hai (1. School of Science Engineering on Communication, Harbin Institute of Technology, Harbin, Heilongjiang 150090, China; 2. Jilin Province Express Way Construction Bureau, Changchun, Jilin 130021, China) Abstract: The present optimal sensor placement is based on single-object criterion in the bridge health monitoring and structure modal test. A new method for optimal sensor placement is presented based on multi-criterion. Using the multi-object optimization algorithm, the satisfying solution (not necessarily optimal in the sense of strict mathematics) of the sensor placement is carried out using Pareto genetic algorithm. The genetic operator and coding precept are designed in this paper. Front side of Pareto solution is obtained at the end of the optimal calculation, proving the feasibility of the proposed method. Key words: modal identification; sensor; optimal placement; genetic algorith