Font Size: a A A

Research Of Freeway Automatic Incident Detection Algorithm Based On Fixed Detector

Posted on:2012-12-23Degree:MasterType:Thesis
Country:ChinaCandidate:X M WeiFull Text:PDF
GTID:2212330338463231Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
The emergence of freeway travel for people and goods has opened up a safe, efficient, comfortable, convenient fast-track, resulting in enormous social and economic benefits. At the end of 2010, the total length of Chinese freeway increase from 41,000 km at the end of Tenth Five-Year to 74,000 km, ranking second in the world, after USA that has about 100,000 kilometers. With the rapid development of freeway, traffic incidents are increasing, and its negative impacts are increasing too. On the one hand, the traffic incidents on the freeway affect larger area than before. The congestion of incidents led to huge economic losses; On the other hand, traffic incidents, and the secondary incidents led to more and more serious casualties. So timely, accurate and reliable detection of traffic incidents is important for traffic management to take measures to quickly restore traffic capacity of the freeway and reduce the losses of life and property.This paper aims to study the characteristics of traffic flow caused by traffic incidents, and study algorithm of Automatic Incident Detection(AID), also study how to improve Identification Rate(IR),and how to reduce False Identification Rate(FIR) and Mean Time To Identification(MTTI) of AID algorithm.The main contents of this paper are as follows.First, a brief introduction of the freeway incident detection research background and status of related technology research and application; describes the basic theory of freeway traffic flow and related concepts, introducing the definition and classification of traffic incidents and the algorithm of AID; Highlights the AID algorithm performance evaluation indicators and evaluation methods; a brief introduction to the several classical algorithms of AID and compare their advantages and disadvantages.Secondly, use VISSIM simulation tools to simulate freeway traffic incident, analyze traffic flow characteristics in different lane closed by incident, different distance between detector and incident sites, different traffic flow. Normal traffic flow characteristics of freeway also discussed.Subsequently, through instance validation shows the standard deviation method's advantages and disadvantages. Analyze the reasons of its advantages and disadvantages, and propose a combinational AID algorithm. In this algorithm, the state of freeway traffic flow is divided into low-flow state and non-low-flow state. When the traffic flow is low-flow state, we use Low-flow AID algorithm. When the traffic flow is non-low-flow state, we use improve standard deviation method. The improve standard deviation method use upstream occupancy and downstream flow as parameters. Compared to the standard deviation method, instance validation shows that the combinational AID algorithm improves the identification rate and reduce the false identification rate, also reduce mean time to detection. Also proposed a new incident remove detection method to further improve the traffic incident detection system.Finally, this article summarizes the research, including innovations and shortcomings, as well as the following improvements.
Keywords/Search Tags:Combinational AID algorithm, Low-flow AID algorithm, Improve Standard deviation method, Incident remove detection method
PDF Full Text Request
Related items