Research On Short-time Prediction Methods Of Traffic Parameters On Urban Road Network

With the continuous development of urban transportation and intelligent transportation technology innovation,urban traffic problem is imminent.Urban traffic induction is an effective strategy to solve the traffic problem of large and medium-sized cities,in the city of induction,the need for predicting the effective,real-time traffic information,and traffic managers can according to the traffic information analysis traffic operation condition of the next moment,and then put forward effective induction planning.Therefore,the role of traffic parameters prediction in traffic management have nots allow to ignore.The traffic problems at present,the first analysis of data sampling interval and xining,find out the proper sampling interval and predict time interval;Was proposed based on grey correlation degree analysis of Kalman filtering algorithm for traffic parameter prediction,forecasting based on historical data,the fitting is good.Considering the part in the city road network detector not covered,or detector damage problems,adopt the method of correlation analysis,no detector for sections of history data acquisition analysis to solve difficult situation.In this paper,from the practical perspective,using a large number of traffic data is analyzed,and to validate the proposed method.Verification results show that the method can effectively solve the problems of the actual traffic parameter prediction,and can provide effective data support for induced traffic.