基于Wi-Fi探针的地铁车站拥堵点客流量估计方法
|
摘要
实现地铁站拥堵位置客流流量的实时监控是提升车站运营管理的关键,也一直是运营安全的难点。Wi-Fi探针对携带Wi-Fi设备的对象进行动态跟踪,实现客流流量的精确化采集。在对Wi-Fi探针的数据采集原理、数据预处理及关联性分析的基础上,通过实测数据建立BP神经网络客流量估计模型,并以上海地铁徐泾东站为例实现模型和算法的验证。结果表明,该技术满足客流流量采集分析的要求,且证明了客流量动态估计方法的可行性和有效性。
Real-time monitoring of passenger flow in congested locations of subway stations is the key to improving the operation and management of stations,and it has always been a problem in operational safety.The Wi-Fi probe dynamically tracked the objects carrying the Wi-Fi device to achieve accurate collection of passenger flow.Based on the data acquisition principle,data preprocessing and correlation analysis of Wi-Fi probe,the BP neural network traffic flow estimation model was established by the measured data.The verification of model and algorithm were carried out by taking Shanghai Xujingdong Station as an example.The results show that the Wi-Fi probe passenger flow acquisition technology meets the requirements of passenger flow collection and analysis.It also proves the feasibility and effectiveness of the passenger flow dynamic estimation.
Real-time monitoring of passenger flow in congested locations of subway stations is the key to improving the operation and management of stations,and it has always been a problem in operational safety.The Wi-Fi probe dynamically tracked the objects carrying the Wi-Fi device to achieve accurate collection of passenger flow.Based on the data acquisition principle,data preprocessing and correlation analysis of Wi-Fi probe,the BP neural network traffic flow estimation model was established by the measured data.The verification of model and algorithm were carried out by taking Shanghai Xujingdong Station as an example.The results show that the Wi-Fi probe passenger flow acquisition technology meets the requirements of passenger flow collection and analysis.It also proves the feasibility and effectiveness of the passenger flow dynamic estimation.
引文
[1]Ling B,Tiwari S,Li Z,et al.A multi-pedestrian detection and counting system using fusion of stereo camera and laser scanner[C]//Spie Optical Engineering+Applications.International Society for Optics and Photonics,2010.
[2]张翔.地铁车站人员人脸图像识别技术研究[J].科技经济市场,2015(10):11.
[3]徐斌,陈晓冰.基于红外传感器的城市客车客流统计系统的研制[J].客车技术与研究,2010,32(05):12-14,17.
[4]Henson C.Levels of sranspervice for pedestrian[J].Transportation Engeering,2000,70(9):26-30.
[5]谢辉,董德存,欧冬秀,等.轨道交通短期客流预测方法及其算法研究[J].现代城市轨道交通,2011(3):96-99.
[6]陈光,段满珍,张林.城市轨道交通网络客流动态优化估计[J].计算机仿真,2016,33(7):210-212,233.
[7]钱雪忠,孔芳.关联规则挖掘中对Apriori算法的研究[J].计算机工程与应用,2008(17):138-140.
[8]杨伟,倪黔东,吴军基.BP神经网络权值初始值与收敛性问题研究[J].电力系统及其自动化学报,2002(1):20-22.
[9]庞梦圆,杜彦君,厉超吉,等.基于图像采集技术的地铁站台客流引导系统研究[J].都市快轨交通,2017,30(2):37-41.
[2]张翔.地铁车站人员人脸图像识别技术研究[J].科技经济市场,2015(10):11.
[3]徐斌,陈晓冰.基于红外传感器的城市客车客流统计系统的研制[J].客车技术与研究,2010,32(05):12-14,17.
[4]Henson C.Levels of sranspervice for pedestrian[J].Transportation Engeering,2000,70(9):26-30.
[5]谢辉,董德存,欧冬秀,等.轨道交通短期客流预测方法及其算法研究[J].现代城市轨道交通,2011(3):96-99.
[6]陈光,段满珍,张林.城市轨道交通网络客流动态优化估计[J].计算机仿真,2016,33(7):210-212,233.
[7]钱雪忠,孔芳.关联规则挖掘中对Apriori算法的研究[J].计算机工程与应用,2008(17):138-140.
[8]杨伟,倪黔东,吴军基.BP神经网络权值初始值与收敛性问题研究[J].电力系统及其自动化学报,2002(1):20-22.
[9]庞梦圆,杜彦君,厉超吉,等.基于图像采集技术的地铁站台客流引导系统研究[J].都市快轨交通,2017,30(2):37-41.