摘要
浮动车技术和线圈检测技术在获取交通数据方面各有所长,为了得到更高质量的数据,需要对GPS浮动车数据与线圈检测数据进行融合。针对此问题,建立了基于BP网络的融合模型,该模型包括三个部分,即初始数据产生模块、BP神经网络数据融合模块和融合结果分析模块。模型的输入参数为线圈检测模块的行程时间、交通量和GPS浮动车检测模块的行程时间、浮动车数量,输出结果为融合后的行程时间。通过仿真试验对融合模型进行了评价,融合后的行程时间的准确度和稳定性都较高,表明数据融合模型是有效的。
The advantages and disadvantages of both floating vehicle and loop techniques are different. In order to get higher quality data, it needs to fuse GPS floating vehicle data with loop detector data. Focusing on this issue, the paper established a fusion model based on BP network. The model consists of three parts, the initial data generated module, data fusion module based on BP network and result analysis module. The input data are travel time, traffic volume provided by the loop detectors and travel time, number of floating vehicle provided by GPS floating vehicle. The output is travel time after fusing. The simulation result shows that the accuracy and stability of the travel time after fusing are high, and the model is effective.
出处
《计算机仿真》
CSCD
北大核心
2009年第9期235-238,共4页
Computer Simulation
基金
国家自然科学基金项目"基于移动与固定检测的路网交通流建模及动态诱导算法研究"(50578064)
关键词
反向传播网络
数据融合
浮动车
行程时间
仿真
Back propagation network
Data fusion
Floating vehicle
Travel time
Simulation