交通流特性参数的识别与测量

Detection with PR Technique for The Character Parameters of Traffic Flow

提出一种用于交通流管理控制系统中交通流特性参数自动检测的新办法 ,介绍该方法的基本原理、识别模型及据此方法研制成功的在线实时智能检测系统的实验结果。这个系统较好地应用了模式识别技术 ,故不仅能自动区分机动车与非机动车 ,判别车型 ,准确地检测出机动车流特性参数 ,而且能自动准确地检测出非机动车 (自行车 )群体流量。此系统样机现场连续运行 3年多 。

For traffic management and city programming,on line real time accurate automatic detection for the character parameters of traffic flow is very important.However,to develop a reliable traffic flow character parameter detecting device(for vehicles and bicycles),as known,is still a knotty problem need to be solved.In this paper,the dynamic difference frequency method with Patterm Recognition(PR)technique by using inductive loop is presented.The basic principle,the recognition models and the testing results of the detecting system for traffic flow character parameters are given.The detecting system based on this method is operating on intersection for more than 3 years and the results are satisfactory.