摘要
基于出行样本数据估计的城市交通起讫点(OD)需求需要进一步扩张到全体出行者数量上,运用数学规划理论研究了检测器布局策略和扩样系数推断方法。考虑路段和路径覆盖信息最大原则,提出检测器布局模型以确定最优的检测器布设数量和位置。根据布设检测器上的观测路段流量,建立扩样系数推断双层规划模型,其中上层目标函数最小化观测流量与待估流量之间的偏差,约束为扩样系数、OD需求和路段流量之间的解析关系,下层采用随机用户均衡分配获取OD-路段关联比例。设计了逐次动态识别检测器和迭代算法分别求解检测器布局与扩样系数推断模型。通过算例表明,整合的检测器布局与扩样系数推断模型估计的扩样系数精度为0.01,建立的检测器布局模型可以用来确定改装检测器的最优策略,设计的算法均可以快速收敛于均衡解。
Since the trip data based traffic origin-destination(OD)demand needs to be expanded to the whole travelers′level counts,the sensor deployment strategy and expansion factor inference are studied using mathematical programming theory.The sensor deployment model is presented to determine the optimal quantity and locations of sensors through considering the principle of maximum both link and route fl ow coverage information.Based on the link flows observed from the deployed sensors,the bi-level expansion factor inference model is built.The objective function of upper-level model minimizes the distances between the observed and estimated traffic flows,and the constraints are the relationships between expansion factor,OD Demand and link flow.The stochastic user equilibrium(SUE)is adopted as the lower-level model to derive the OD-link proportions.The sequential identifying sensor location algorithm and iterative algorithm are designed to solve the sensor deployment strategy and expansion factor inference model,respectively.Numerical examples demonstrate that the accuracy of values estimated by integrating sensor deployment strategy and expansion factor inference model can reach to 0.01;the built sensor deployment strategy can also be used to determine the optimal scheme of refitting sensors;and the designed algorithms can make convergence to the equilibrium solutions rapidly.This research has significant promoting effects on developing the theory of mobile phone data based OD demand estimations.
作者
孙超
尹浩为
汤文蕴
褚昭明
SUN Chao;YIN Hao-wei;TANG Wen-yun;CHU Zhao-ming(School of Automotive and Traffic Engineering,Jiangsu University,Zhenjiang 212013,China;College of Automobile and Traffic Engineering,Nanjing Forestry University,Nanjing 210037,China;Research Institute for Road Safety of MPS,Beijing 100062,China)
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2023年第4期1070-1077,共8页
Journal of Jilin University:Engineering and Technology Edition
基金
国家自然科学基金项目(71801115)
教育部人文社会科学研究基金项目(22YJCZH153)
江苏省研究生科研与实践创新计划项目(SJCX22_1877)
公安部科技计划项目(2022LL81)。
关键词
交通运输系统工程
检测器布局
扩样系数推断
逐次识别
迭代算法
engineering of communication and transportation system
sensor deployment
expansion factor inference
sequential identifying
iterative algorithm
