基于卡口监测数据的路段交通异常状态识别方法研究  

Research on Recognition Method for Abnormal Traffic State of Road Section Based on Checkpoint Monitoring Data

作　　者：张雷元 张韧 刘海平 ZHANG Leiyuan;ZHANG Ren;LIU Haiping(Traffic Management Research Institute of the Ministry of Public Security,Wuxi 214151,China;Traffic Police Corps,Public Security Bureau of Xinjiang Uygur Autonomous Region,Urumqi,830011,China)

机构地区：[1]公安部交通管理科学研究所,江苏无锡214151 [2]新疆维吾尔自治区公安厅交通警察总队,新疆乌鲁木齐830011

出　　处：《中国人民公安大学学报（自然科学版）》2023年第1期83-87,共5页Journal of People’s Public Security University of China(Science and Technology)

摘　　要：以路段为对象,研究交通异常状况自动识别方法,为道路交通管理部门提供技术手段。首先,系统分析了现有交通异常状态识别方法原理和存在的问题,结合卡口监测数据特性,设计异常状态识别的基本思路及研究路线;其次,对路段模型进行重新定义,利用其上下游卡口历史数据,获取行程时间指标参数;再次,基于行程时间概率分布规律,确定异常状态对应的行程时间阈值,建立识别流程,将实测行程时间与行程时间阈值进行比较,实现异常状态判别;最后,选取乌鲁木齐市建国路部分路段开展实例验证。结果显示,采用所提方法进行路段异常状态评价,其评价结果与人工观测情况完全相符,能有效提升异常状态识别的实时性、准确性。Taking the road section as the research object,the recognition method for abnormal traffic state was investigated,which provided technical means for road traffic management departments.Firstly,the existing methods for abnormal state recognition and the corresponding problems were systematically analyzed,and the basic idea and research route of abnormal state recognition were designed according to the characteristics of the monitoring data at the checkpoints.Secondly,the road section model was redefined.By using the historical data of upstream and downstream checkpoints,travel time indexes were obtained.Thirdly,based on probability distribution of travel time,the travel time threshold corresponding to the abnormal state was determined,and identification process was also proposed.The measured travel time was compared with the travel time threshold to realize the abnormal state discrimination.Finally,several road sections of Jianguo Road in Urumqi were selected for verification.The results showed that the evaluation of the proposed method was completely consistent with the manual observation,which could effectively improve the real-time and accuracy of abnormal state recognition.

关 键 词：交通管理工程 异常状态识别方法 行程时间分布规律 卡口监测数据 

分 类 号：D035.37[政治法律—政治学]