城市交通控制中的数据采集研究综述

过去几十年来，城市交通控制技术为适应不断增长的交通需求和日益复杂的管理目标有了长足发展. 作为城市交通控制策略制定和控制算法设计的基础，交通数据决定了城市交通控制的适用性、可靠性和先进性. 数据采集技术的发展为城市交通控制能力的提升带来了机遇和挑战. 本文回顾交通控制系统中数据采集与参数估计的基本方法，分析评述检测数据方法从固定式无标识数据、移动式检测数据到固定式有标识数据的演变，指出它们给交通参数估计带来的变革. 结合20 世纪末出现的移动式检测技术，分析评述了对应的两种交通参数估计方法，即基于概率论的方法和基于交通流激波理论的方法给交通控制带来的影响. 针对近年来出现的固定式有标识检测数据，分析其对城市交通需求估计及交通控制策略参数估计研究带来的新任务. 最后，分析指出我国未来交通控制研究的三个方向：一是城市交通控制系统的信息范围已扩展到区域和路网层级，二是交通参数估计研究的重点随着数据采集方式的演变已转向提高参数估计的实时性与精度等领域，三是交通参数研究理论与实践存在差异，如何结合我国城市交通系统运行中检测环境差异导致数据误差率的变化，交通流到达规律的变化，道路上不同种类交通流间的交叉干扰等实际应用因素，使方法与模型能有效地指导我国复杂的交通控制实践是重要方向.

Review of Traffic Data Collection Research on Urban Traffic Control

Over the past few decades, urban traffic control technology has evolved to meet growing traffic demands and increasingly complex management objectives. As the basis of urban traffic control strategy and control algorithm, traffic data determines the applicability, reliability and advancedness of urban traffic control system. The development of data collection technology has brought opportunities and challenges to the improvement of urban traffic control. This paper reviews the basic methods of data collection and parameter estimation in traffic control systems. It analyzes the evolution of detection data methods from fixed unmarked detection data methods, probe vehicle- based data to fixed unique data. Combined with the probe vehicle- based data that emerged at the end of the 20th century, the impact of two corresponding traffic parameter estimation methods (stochastic method and shockwave-based method) are analyzed and reviewed. In view of the fixed unique detection data that has appeared in recent years, this paper analyzes the new tasks on urban traffic demand estimation and parameter estimation in traffic control. The paper then points out three directions of future trafficcontrol research in China: first, the information scope of urban traffic has been extended to regional and road network levels; second, with the evolution of data collection methods, the focus of traffic parameter estimation research has shifted to improving the real-time and accuracy of parameter estimation; third, there are gaps between the theory and practice in traffic parameter estimation. It is an important direction in urban traffic control research on how to use those methods and models to better guide the complex traffic control practice in China and consider the data error caused by the detection under different urban transportation environment, traffic flow arrival rules, and cross-interference between different types of traffic flows.