基于BP神经网络的桥梁多传感器信息融合的研究  

Bridge Multisensor Information Fusion Research Based on BP Neural Network

作　　者：沈健[1] 钱国明[1] 谢阳阳[2] 

机构地区：[1]南京邮电大学电子科学与工程学院,江苏南京210003 [2]南京信息工程大学,江苏南京210044

出　　处：《仪表技术与传感器》2017年第3期110-112,117,共4页Instrument Technique and Sensor

摘　　要：在现代桥梁多传感器健康监测系统基础上,建立了基于BP神经网络的桥梁多传感器信息融合系统。首先对桥梁不同断面的倾角传感器采集的样本数据进行特征提取,并将特征值进行关联组成特征序列,利用BP神经网络在模式识别问题上的优势,对特征序列进行信息特征层融合。通过仿真实验,设计了可用于一般桥梁数据模式识别问题的3层BP神经网络,且具有准确率高、收敛快的特点。On the basis of modern bridge health monitoring system for many sensors,bridge multisensor information fusion system based on BP Neural Network was established.Through the obliquity sensors to collect sample data in feature extraction of bridges of different section,and characteristic sequence was associated using characteristic.Advantages of the BP neural network in the field of pattern recognition were used,and signature sequence was fused on the layer feature.Through the simulation experiment,a three-layer BP neural network that can be generally used in the pattern recognition problem of bridge data was designed,which has characteristics of high accuracy and fast convergence.

关 键 词：信息融合 BP神经网络 模式识别 特征层融合 

分 类 号：TH7[机械工程—仪器科学与技术]