摘要
针对当前桥梁结构健康状态检测模型存在的检测误差大、检测效率差等难题,以改善桥梁结构健康状态检测结果为目标,设计了基于数据挖掘技术的桥梁结构健康状态检测模型。首先,采用无线传感器网络对桥梁结构健康状态数据进行采集,并采用核主成分分析对桥梁结构健康数据处理,去除桥梁结构健康数据的冗余特征,减少桥梁结构健康状态检测特征规模;然后,采用支持向量机对桥梁结构健康数据进行学习,并引入粒子群优化算法确定的桥梁结构健康状态检测模型的参数,建立最优的桥梁结构健康状态检测模型;最后,在Matlab2017平台上对桥梁结构健康模型的有效性和优越性进行测试。结果表明:本文模型获得了较高精度的桥梁结构健康状态检测结果,桥梁结构健康建模时间减少,提高了桥梁结构健康状态检测效率,而且桥梁结构健康状态检测整体性能要明显优于当前其他的桥梁结构健康检测模型,为桥梁结构健康研究提供了一种有效的工具。
In order to improve the health detection results of bridge structures,a health detection model of bridge structures based on data mining technology is designed to solve the problems of large error and poor detection efficiency in the current health detection model. Firstly,the health data of bridge structure is collected by wireless sensor network,and the data is processed by kernel principal component analysis,which eliminates the redundancy of bridge structure health data and reduces the detection scale of bridge structure health. Then,the support vector machine is used to learn the health data of bridge structure. The parameters of bridge structure health detection model determined by particle swarm optimization algorithm are introduced to establish the optimal bridge structure health detection model. Finally,the effectiveness and superiority of the bridge structure health model are tested on the platform of Matlab 2017. The results show that the model achieves high precision bridge structure health. The health detection results show that the time of bridge structure health modeling is reduced,and the efficiency of bridge structure health detection is improved. Moreover,the overall performance of the proposed bridge health detection model is obviously better than that of other bridge health detection models. The proposed model provides an effective tool for bridge structure health research.
作者
武立群
张亮亮
WU Li-qun;ZHANG Liang-liang(CET-College of Engineering and Technology Southwest University,Chongqing 400715,China;School of Civil Engineering,Chongqing University,Chongqing 400045,China)
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2020年第2期565-571,共7页
Journal of Jilin University:Engineering and Technology Edition
基金
国家自然科学基金项目(50778185).
关键词
桥梁结构
健康状态
核主成分分析
冗余特性
检测精度
建模效率
bridge structure
health status
nuclear principal component analysis
redundancy characteristics
detection accuracy
modeling efficiency
