基于多源数据时点匹配的光电测量信息深度融合方法  

Depthfusion method of photoelectric measurement information based on time point matching of multi-source data

作　　者：孙彤 刘志萍[1] 褚俊英[1] 李孟娇 SUN Tong;LIU Zhiping;CHU Junying;LI Mengjiao(Haibin college,Beijing Jiaotong University,Huanghua Hebei 061199,China)

机构地区：[1]北京交通大学海滨学院,河北黄骅061199

出　　处：《激光杂志》2021年第8期93-97,共5页Laser Journal

基　　金：河北省高等学校科学研究项目(No.QN2020523);沧州市重点研发计划指导项目(No.192107004);河北省高等教育教学改革研究与实践项目(No.2019GJJG640)。

摘　　要：由于缺乏对时间特征的考虑,导致光电测量信息融合程度不够,即存在信息覆盖率低、信息丢失率大的问题。针对上述问题,从时点匹配角度着手,对多源光电测量信息进行深度融合研究。研究分为三部分:先对采集到的光电测量信息进行去噪和标准化预处理,然后从处理好的信息中利用核主成分分析法提取特征,并表示成基准时间序列形式,最后进行相关性判定,实现时点特征匹配,完成光电测量信息深度融合。结果表明:与基于模糊理论、D-S证据理论、神经网络的三种信息融合方法相比,本方法信息覆盖率高、信息丢失率小,证明本方法的信息融合质量更高。Due to the lack of consideration of time characteristics,the degree of information fusion of photoelectric measurement is not enough,that is,the problem of low information coverage and large information loss rate.Aiming at the above problems,this paper studies the depth fusion of multi-source photoelectric measurement information from the point of time matching.The research is divided into three parts:firstly,denoise and standardize the collected photoe-lectric measurement information,then extract the features from the processed information by kernel principal compo-nent analysis,and express them into the form of benchmark time series.Finally,the correlation is determined to realize the time point feature matching and complete the depth fusion of photoelectric measurement information.The results show that:compared with the three information fusion methods based on fuzzy theory,D-S evidence theory and neural network,this method has higher information coverage rate and lower information loss rate,which proves that the infor-mation fusion quality of this method is higher.

关 键 词：多源数据 时点匹配 光电测量信息 融合方法 

分 类 号：TN249[电子电信—物理电子学]