Prediction Study For Urban Link Travel Time Based On Floating-car Data

With the continuing worsening of urban traffic situation, people have a more intense demand for the real-time, accurate, reliable travel time information. Floating-car system is a kind of newly developed traffic flow information collecting technology with ITS new technology's applications these years , it could be a helpful supplement to traditional methods. The travel time prediction based on floating-car data is an important research item in the field of Intelligent Transportation Systems. It can provide real-time, accurate, reliable travel time prediction information for its core subsystems Advanced Traveler Information System and Advanced Transport Management Systems of the ITS, and satisfy what the system function need for travel time prediction, so that ITS can quickly and accurately find out road network operation state, then make the benefit of the society and the economy biggest.Therefore, this paper focuses on the prediction study for urban link travel time based on floating-car, try to get an accurate prediction for it. First, this paper introduces the new floating-car traffic information collecting system, including its definition,principles and the process of getting travel time data. Then on the basis of the actual travel time data, paper carries on the link travel time forecast by BP artificial neural network based on Bayesian algorithm. Then we established many prediction examples to carry on the confirmation to this method according to the different network architecture and sample data. Finally this paper researches the possibility of further increases for prediction precision by Genetic Algorithm- BP artificial neural network.The experiments prove that the result of BP artificial neural network prediction for urban link travel time based on floating-car data can satisfy the precision purpose well. And the Genetic Algorithm-BP artificial neural network precision has a further enhancement.