基于数据融合的交通流量数据在线校正  

On-line calibration of traffic flow data based on data fusion

作　　者：韦达利[1] 陈锋[1] 卞凯[1] 令狐彬[1] 

机构地区：[1]中国科学技术大学自动化系,安徽合肥230027

出　　处：《中南大学学报（自然科学版）》2009年第S1期341-346,共6页Journal of Central South University:Science and Technology

基　　金：国家高技术研究发展计划项目(2007AA11Z222)

摘　　要：针对传感器数据采集过程中出现的数据缺失或异常,通过对相邻车道交通流量数据的相关性的研究与分析,提出一种交通流量数据的在线校正模型,该模型采用最小二乘支持向量机回归融合相邻车道交通流量数据,并结合历史同期数据对交通流量数据的异常进行在线判断与校正。该模型应用于现场采集交通流量数据的校正处理,实验结果表明该模型的有效性。In order to repair missing data and abnormal data, according to analysis of the correlation among adjacent lanes, an on-line calibration model for traffic flow data was proposed based on data fusion. The least squares-support vector regression (LS-SVR) was introduced to fuse the adjacent lanes’ traffic flow data in this model. Meanwhile, the past traffic flow data were integrated with predicted results to identify and calibrate abnormal data. This model was used to process the field data. The experimental results indicate that this model is effective and practical.

关 键 词：交通流量数据 在线校正 数据融合 最小二乘支持向量机回归 

分 类 号：U491.113[交通运输工程—交通运输规划与管理]