基于云计算的卡尔曼滤波速度诱导模型研究  

A Speed Guidance Model of Kalman Filter Based on Cloud Computing

作　　者：张丽岩[1,2] 马健[1] 李克平[1] 孙焰[1] 

机构地区：[1]同济大学道路与交通工程教育部重点实验室,同济大学交通学院,上海201804 [2]苏州科技学院土木工程学院,江苏苏州215011

出　　处：《公路工程》2013年第6期92-96,105,共6页Highway Engineering

基　　金：国家自然科学基金(61004113和71072027);江苏省高校自然科学基金(12KJB580005)

摘　　要：利用自适应卡尔曼滤波器来建立城市快速路瞬时交通预测模型,同时通过Hadoop基础框架及MapReduce的编程模型设计和开发了交通属性指标最小约简云处理系统,并在此基础上对车辆进行速度诱导控制,建立了以车辆总行程时间为目标函数的诱导优化模型。并以上海南北高架南段东线为例建立了仿真场景,对模型进行了参数标定与校验,通过有效性与性能分析表明:用云计算能够提高运算效率,且速度诱导控制后,降低了路段上交通流的整体速度差异,提高了交通安全性。The paper establishes the speed guidance model of urban expressway based on the Kal- man filter, which can provide the role of short-term traffic flow prediction. And the paper designs and de- velops the minimum attribute reduction system of the transport properties indicators based on Hadoop and MapReduce. In addition, the model regards gross vehicle travel time as the objective function to proceed optimized design. Then, it simulates and analyses the results based on the east line of the southern sec- tion of Shanghai North-South Elevated road. Simulation result shows that cloud computing can improve the operation efficiency. And the speed guidance control reduces the overall difference of speed of the traffic flow that can improve traffic safety and the model is effective.

关 键 词：云计算 HADOOP MAPREDUCE 卡尔曼滤波 速度诱导 最小约简 

分 类 号：U491.112[交通运输工程—交通运输规划与管理]