摘要
为实现桥梁加速度传感器的优化布置,提出一种基于神经网络和遗传算法的布置方法。利用ANSYS软件建立桥梁模型并获得模态数据,通过随机生成大量布置方案及对应MAC值形成数据集,并建立双隐藏层的神经网络模型进行训练,将训练好的模型利用遗传算法搜索出最优值及对应的布置方案,最后对结果进行了分析,表明该方法可行。
As for bridge acceleration sensor,a layout method is presented based on neural network and genetic algorithm.By using ANSYS software,bridge model was established and the modal data was obtained.A data set was formed with randomly generated layout scheme and the corresponding MAC value.Neural network model of double hidden layer was set up to train.The trained model was used to search the optimal value and the corresponding layout scheme based on genetic algorithm.Experiment results show that the method is feasible.
作者
袁灿
唐川田
李文钊
YUAN Can;TANG Chuantian;LI Wenzhao(School of Information Science and Engineering,Chongqing Jiaotong University,Chongqing 400074 China;School of Information Science and Technology,Chengdu University of Technology,Chengdu 610059 China)
出处
《西华大学学报（自然科学版）》
CAS
2018年第2期13-18,共6页
Journal of Xihua University:Natural Science Edition
基金
重庆市科委基金项目(cstc2016shmszx30026)
重庆高校创新团队建设计划项目(CXTDG201602013)
乌鲁木齐市科学技术计划项目(Y161320008)
关键词
传感器
优化布置
神经网络
遗传算法
sensors
optimal placement
neural networks
genetic algorithm
