Jam Evaluation And Short-term Forecasting Of Urban Expressway Based On Weather

With the development of the city,urban traffic congestion has become an increasingly common phenomenon.The peak of commuting,accident,rain and snow weather and many other factors will cause jam.Road congestion not only makes the residents travel inconvenience,but also for the traffic management department brought great trouble.With the extensive application of intelligent transportation technology,collection traffic volume,velocity and delay is more convenient.And provides data and technical support for traffic congestion evaluation.In recent years,congestion indentification,congestion evaluation and congestion prediction become one of the more active areas of transportation field.But so far very few experts and scholars do traffic congestion evaluation and prediction combining with rain,snow and other weather factors on.Rain and snow weather often occurs in the city.Rain and snow weather often cause different degrees of traffic congestion.The paper aims to combine the weather factors to develop accurate and effective evaluation and short-term forecasting model for expressway.First of all,collecting traffic volume,velocity and delay,analyzing the change of traffic volume,velocity and delay in the role of precipitation and other climate parameters.Doing correlation analysis using SPSS.And we know the change of traffic characteristics in rain and snow weather.Secondly,using BP neural network as tool,using v/c,velocity and delay as evaluation index,and then we evaluate the degree of traffic congestion.Thus we get a congestion index between 0 and 1.The more close to 0,the more smooth traffic,the more close to 1,traffic congestion.Analyze the relationship between precipitation and congestion index in rain and snow weather.At last,using 3(rd)order morkov model do congestion short-term forecasting.Calculating state transition probability matrix,and forecasting the short-term congestion on Xi'an Second Ring Road in rain and snow weather,and comparing with practical jam grade,thus we can modified 3(rd)order morkov model.In the case analysis,choosing Xi'an Second Ring Road's traffic parameters in rain and snow weather to test the model,and then comparing with the practical jam grade.The predictive accuracy in rain is 82.46%.The predictive accuracy in snow is 77.19%.It Proves the model is effective.