广东水利水电Issue(9)：24-30,7.
基于多方法融合的流量在线自动监测研究
Research on Online Automatic Monitoring of Traffic Based on Multi Method Fusion
摘要
Abstract
In order to effectively improve the accuracy of online automatic flow monitoring at Wuzhou Hydrological Station based on multi method fusion,this study attempts to integrate the monitoring values of different flow monitoring methods,fully leverage the advantages of different flow monitoring methods,and analyze the applicability of different fusion methods in the Wuzhou Hydrological Station.Firstly,the monitoring values of a single traffic monitoring method are fused separately using information entropy,BP neural network and Support Vector Machine,and traffic fusion models based on information entropy,BP neural network and Support Vector Machine are established;Next,the mean absolute error(MAE),root mean square error(RMSE),coefficient of certainty(DC),relative error(RE),and qualification rate(QR)are introduced to evaluate the flow fusion effect and study its applicability in the Wuzhou hydrological station of the Xijiang Golden Waterway.The results indicate that both the fusion traffic evaluation indicators based on information entropy,BP neural networkand Support Vector Machine are superior to single traffic monitoring methods;Traffic fusion can expand the optimal application range of traffic automatic monitoring methods to indicator flow rates greater than 0.2m/s and stable;Compared with the fusion traffic based on information entropy and Support Vector Machine,the fusion traffic based on BP neural network has the highest accuracy.关键词
信息熵/BP神经网络/支持向量机/流量在线自动监测/流量融合/西江Key words
Information entropy/BP neural network/Support Vector Machine/Online automatic monitoring of traffic/Traffic fusion/Xijiang River分类
天文与地球科学引用本文复制引用
朱颖洁..基于多方法融合的流量在线自动监测研究[J].广东水利水电,2025,(9):24-30,7.基金项目
国家自然科学基金项目(编号:41461005) （编号:41461005）
广西自然科学基金重点项目(编号:2022GXNSFDA080009) （编号:2022GXNSFDA080009）
广西水利厅科技项目(编号:201618). （编号:201618）
