Application Of Date Fusion Technique In Freeway Traffic Incident Detection

In recent decade years, the advanced development of highway provides a safe, efficient, comfort and convenient way for travels of people and goods, but at the same time, it also put forward higher requirement for management, control and the traffic detection of highway. Hence the study on how to effectively use incident management technologies to enhance the operation on freeways is a critical research field in traffic engineering.There are several kinds of incident detection algorithms of freeway which have perfect result in perspective, but there haven't any algorithm used in practice. The reasons are that the traffic data that collected by the fixed detectors and the probe vehicles is not effective enough for incident detection algorithm because of the space collocation of the fixed detectors is too large, or the sample of the probe vehicles is too small. If you want to develop an incident detection capability, you should augment density of the fixed detectors or sample of the probe vehicles. But it can't realize in recently years.This dissertation based on"The Study on the Theory and Methodology of Freeway intelligent Incident Surveillance and Control System". Under this background, the thesis performs the application of data fusion technique in incident detection of freeway.The thesis first introduced two incident detection algorithms of freeway which are incident detection algorithm of freeway based on date layer fusion and incident detection algorithm of freeway based on decision layer fusion to detection different severity incidents occurred on different location on link. The purpose is to validate pertinence of the capability of incident detection algorithms and the sample of the probe vehicles. Then introduced different combination of traffic detectors and test the capability of incident detection algorithms in different conditions.This thesis not only consider traffic incident of different location and severity, analyze the infection of the space collocation of the fixed detectors and the sample of the probe vehicles to incident detection algorithms, but also use for reference three level reports to police system to satisfy to different users. The thesis comprises of 6 chapters and their contents are as follows:Chapter One: Introduction. First, explain what and why to be investigated in the thesis. Review the history and present situation of incident detection algorithm of freeway all over the world. After that, the main content and structure of the thesis is given.Chapter Two: The elements of study. Introduce the theory, levels and methods of date fusion. After that, analyze the feasibility of the application of date fusion in incident detection algorithm of freeway.Chapter Three: The study of incident detection algorithm of freeway based on date layer fusion. The chapter project incident detection algorithms of freeway based on date layer fusion, and analyze the feasibility of PNN used in incident detection algorithm. Then the theory and steps of the incident detection algorithm are given.Chapter Four: The study of incident detection algorithm of freeway based on decision layer fusion. The chapter project incident detection algorithms of freeway based on decision layer fusion, and the theory, steps of the incident detection algorithm and the algorithms which are fused are given.Chapter Five: The study of combination optimization of traffic detectors. The chapter gives the purport, theory and steps of the optimization of the space of the fixed detectors and the sample of the probe vehicles. Then the chapter gives different traffic detector combined projects, and the purport, theory and steps of the optimization are given.Chapter Six: Analysis using simulated date. The chapter simulates different incidents of freeway, and analyzes incident detection algorithm based on date layer fusion, incident detection algorithm based on decision layer fusion and combination optimization of traffic detectors using simulated date. The result is that if we use the traffic date collect by fixed detectors and probe vehicles, the space of the fixed detectors is bigger than incident detection algorithm based on fixed detector and the sample of the probe vehicles is smaller than incident detection algorithm based on probe vehicles.Chapter Seven: Conclusions and the future work. The thesis is summed up and the main contents of the thesis are looked back. At the same time, point out the problems that need to further study in the future.