Highway Traffic Congestion Detection And Travel Time Prediction Based On Microwave Data

With the increasing number of vehicle in our country, road congestion, traffic accidents and other issues become more serious. Traffic congestion and travel time have become the main concern of the public. How to make timely and accurate traffic congestion detection and travel time prediction has become one of the hot spots in the research of modern intelligent transportation system (ITS). Therefore, the traffic congestion detection and travel time prediction algorithm were researched in this paper based on microwave data from Beijing section of Beijing-Harbin Expressway. The main work and achievements are as follows:(1) The advantages and disadvantages of the basic data acquired from current highways were studied, and then the microwave data was selected as basic research data and the traffic flow state changing characteristic was analyzed. Aiming at the error data of the microwave data, a kind of error data identification method based on threshold value method and traffic flow theory was put forward; then Aiming at the lane detection characteristic of the microwave detectors on the Expressway, a king of data integration strategy was proposed. Using the original traffic flow data as an example, the proposed method is verified effective and feasible.(2) The basic principle of traffic congestion detection was studied at first. Considering the nonlinear change characteristics of traffic flow and detection range limitations of single detector, a kind of traffic congestion detection algorithm based on pattern matching and multi-node correlation analysis was put forward. Through historical traffic congestion pattern library establishment, single node mode matching and multi node correlation analysis, the detection of traffic congestion was completed. Actual application showed that the method based on pattern matching and correlation analysis can automaticly and real-time detect traffic congestion.(3) The advantages and disadvantages of different travel time prediction methods are studied and analyzed. Combined with the characteristics of microwave data, a kind of travel time prediction algorithm based on pattern matching method was proposed. In the method, speed, traffic volume, period of time, road section and traffic congestion level were selected as state parameters to establish a historical pattern library Aiming at the sparse problem of microwave data, a two-dimensional linear interpolation and piecewise method was applied to calculate the travel time between toll stations. The Euclidean distance is used as the criterion to search the similar model by K nearest neighbor strategy, and the weighted average forecast function is used to predict the travel time. Actual application on the test section showed that the pattern matching method can accurately predict the travel time in different traffic flow states.(4) The traffic congestion detection and travel time prediction system was designed and developed. Based on the analysis of the system functional requirements, the overall logical framework of system was designed. Then, the system was implemented based on VS2010, SQL Sever2008 platform, including the system database design, the traffic congestion detection algorithm programming design, travel time prediction programming design and results show platform development.