Recently, the services which provide traffic information to the drivers by detecting traffic situation, delays, and congestions by using various information and communication technologies in the traffic field have increased. The traffic information is...
Recently, the services which provide traffic information to the drivers by detecting traffic situation, delays, and congestions by using various information and communication technologies in the traffic field have increased. The traffic information is providing diverse conveniences such as informing the drivers of the traffic situation of certain sections and of detours in case of congestions. The Smart Highway detects risk factors such as freezing or droppings on the front road by using radars to allow drivers to drive safely under bad weather such as fog or rain or at night and provides the drivers with the information. The relevant researches so far mostly have detected the unexpected situations by collectively processing the massive video data and speed data transmitted to the traffic center from many field systems such as image detection system and loop detection system at the traffic center and provided information according to the manager's judgment finally. However, it has a few limits before being applied to the intelligent highways of the future like the Smart Highway. First, is the technological limit in terms of processing the massive information incoming to the center from the field system and the depreciated sense of real-time of the information. Second, since the collected traffic information is determined based on the experience of the manager, it lacks an automated system for determining the road conditions. Third, since the current traffic information service is limited to providing information on traffic flow and informing of the detours when delayed, it is unable to provide services for prevention of accidents of high-speed drivers and for safe driving. Thus, this doctoral dissertation proposes the following three methods for a real-time road Situation detection model which allows an immediate provision of information to the high-speed drivers by detecting the sudden changes of road conditions on the front. First, study methods to analyze road condition changes based on complex event processing to promptly process the massive road condition data. Second, study ways to measure the similarity of the collected road condition data to the standard data of road condition to predict the road condition real-time and to provide the information promptly. Third, study methods to automatically determine the road condition with the highest probability based on the predicted road condition data of a certain period. This doctoral dissertation collects the road condition data through road radars on the actual experiment road, implements road condition detection system, and conducts a few experiment scenarios to verify that the proposed model is effective via field experiment. The proposed model has displayed excellence in road condition detection performance, accuracy, and functionality in the perspectives of drivers, road managers, and users of the Smart Highway.