The Design And Implementation Of Short-Term Traffic Flow Forecasting Server

It is profound for the research of short-term traffic flow forecasting in intelligent transportation system.Short-term traffic flow forecasting system can not only make travelers effectively avoid the traffic rush to save travel time, but also reduce the auto emissions caused of vehicles running at low speeds.In that case, it is also propitious to reduce the environmental impact of automobile exhaust.This thesis is based on Windows. This system adopted I/O completion port to complete the data receiving module, used the SQL Server stored procedure and spatial database to make the GPS coordinates matched with the roads and forecasted short-term traffic flow by the Kalman filter, finally published the short-term traffic flow forecasting result on the Web platform. This thesis uses actual data as a basis for design, so as to have certain feasibility.According to above design, this thesis completed the following work:(1) Realized the data receiving module.Compared the advantages and disadvantages of I/O completion port server with the common multithreading server, and finally adopted the I/O completion port to design the data receiving module and tested it’s performance.(2) Accomplished the map matching algorithm with a point to a line method. Imported the GIS road layer into SQL Server2008database, then, designed the map matching algorithm with stored procedure and spatial database of SQL Server.(3) Forecasted short-term traffic flow with Kalman filter by the method of single model and single measurement. Only need to update the table in database, which is designed to save the Kalman fiter result, can realize short-term traffic flow forecasting.(4) Released the short-term traffic flow to the Web platform. Clients can get the Json data which is formed by the short-term traffic flow predition data through the URL.