Study On The Methods Of Located Traffic Parameters Short-term And Multi-steps Prediction For Expressway

Traffic congestion is increasingly severe and affects greatly the travel of people. It brings serious air pollution, time delay, energy consumption, and traffic accidents. The problems of traffic jam are urgent to be solved over the world.Major traffic jams governance has static and dynamic methods. The static managements include expansion and reconstruction of roads, intersection canalization, and traffic signal installation and so on; dynamic methods based on real-time road traffic conditions to make traffic management and traffic control in order to alleviate the traffic jam.. Static methods are adopted to meet road traffic demand by improving capacity of roads and reducing traffic conflicts to ease traffic jams, their effectiveness are long, but need to use certain land resource. Generally, traffic congestion does not exist during the whole day on all roads, while the road network have been built, the dynamic managements can make full use of the road resources and reduce traffic congestion. Therefore, with limited road traffic resources, dynamic managements are economical and effective method to solve traffic jams.Dynamic traffic methods rely on traffic information acquisition, and the result of dynamic traffic management depends on the quantity and quality of dynamic traffic information. This information includes real-time and forecast information about traffic condition. We can get road traffic congestion promptly and make management to solve it by analyzing real-time traffic data; We also know where the congestions will happen by analyzing future transportation information and to make management to avoid it. The second management is more effective than the first. However, as a result of information transmission, the real-time traffic data has time delay, so learning more about current and future traffic information need short-term traffic-parameters prediction.The forecast information of traffic parameters has the different application, the specific needs of the different application are not same. These different demands include the forecast time scale, as well as the forecast time length, namely in the forecast time scale definite situation, the number of step needs to forecast. Traffic control system need the recent short-term to make next time traffic control plan; Traffic flow guidance, traffic management and traffic state recognition need the time scale of traffic-parameters prediction is longer than traffic control system; Traveler would like to obtain a longer period of traffic flow forecasting information. Therefore, the traffic parameters of short-term and multistep prediction are very useful.Urban Expressway is an important intra-city and inter-city traffic hub. It bear the mass rapid and long-distance transit, and it plays an important role at traffic Gathering and dispersing. Urban expressway is the key point of urban traffic management. Therefore, Study on the methods of different time scale traffic parameters of short-term and multisteps prediction for urban expressway is very important.This paper studies on the methods of different time scale traffic parameters of short-term and multi-steps prediction for urban expressway, based on This dissertation based on the National High-tech Research and Development Program (863) project (2007AA11Z245).This paper comprises of 6 chapters, and their contents are as follows:Chapter 1: Introduction. Introduce the project origin and explain what and why to be investigated in this paper. Point out the value of this paper both from theory and practice, and then summarize the research status both at home and abroad. At last, the main contents and structure of this paper are given.Chapter 2: Study on the data synthesis of traffic parameters of different time scales, and analysis of short-scale transport parameters of the characteristics of time-series data.Chapter 3: Long-term trend forecast for short-term traffic parameters. According to the transportation parameter data's characteristic of short-term, we choose the Prediction model, then, carries on the empirical analysis to the forecast model.Chapter 4: Metaphase trend forecast for short-term traffic parameters. According to the transportation parameter data's characteristic of short-term, using the improvement adaptive weight exponential smoothing, the improvement grey prediction model, as well as the multi-model fusion algorithm to complete forecast. And then, carry on the empirical analysis to the forecast model.Chapter 5: short-term catastrophe prediction for traffic parameters. In the chapter,we design forecast technique and carry on the empirical analysis to the forecast model.Chapter 6: Concludes the whole paper, summaries the findings and achievements, and brings forward the issues for further research.