Research On The Congestion Prediction And Classification For Expressway Traffics Using Travel Planes

In recent years,with the increase in people’s travel demand and the rapid growth of car ownership,urban road congestion has become more and more serious.At the same time,smart transportation continues to develop,and driverless technology is becoming more mature.Realizing the transparency of the road ahead,improving the smart traffic management system,and alleviating traffic pressure have become the primary tasks of transportation development.In order to realize the transparency of the road ahead,allow autonomous vehicles to enter people’s lives,and provide the possibility of man-machine mixed running,the author’s scientific research team put forward the idea of travel planning,and is committed to building a smart traffic management system based on travel planning.At present,urban road congestion is serious.How to improve the smart traffic management system based on travel plans and relieve traffic pressure is the goal of my research team at this stage.In order to alleviate traffic pressure,this paper introduces the idea of travel planning,and proposes a mechanism for expressway congestion prediction and congestion type discrimination,using travel planning data,real-time and historical traffic data to achieve congestion prediction,congestion type discrimination and rapid response to traffic incidents.Aiming at the two problems of congestion prediction and congestion type discrimination,this paper proposes an RBF neural network speed prediction model based on travel planning and time-space characteristics and a three-dimensional McMaster algorithm congestion type discrimination model based on cusp catastrophe theory.The main research contents are as follows:(1)The congestion prediction based on travel plan and the feedback mechanism of congestion type discrimination are proposed.This paper applies travel planning to traffic congestion prediction and congestion type discrimination,and proposes a congestion prediction and congestion type discrimination feedback mechanism.The feedback mechanism obtains travel plan data in real time from the cloud computing service platform,uses data collection and collection technology to obtain road network real-time traffic flow data,obtains historical data from historical databases,and inputs speed data into the RBF neural network for speed prediction.To achieve the purpose of road congestion prediction.The congestion prediction results are stored in the congestion database.When the predicted time point is reached,the actual congestion situation is compared with the forecast situation,the unpredicted congestion is screened out and the traffic flow data is input into the three-dimensional McMaster algorithm based on the cusp catastrophe theory.Type discrimination model.The type of congestion will be fed back to the traffic control department and travel users through the wireless base station in time.If it is an occasional traffic jam,it can help traffic control departments to quickly deal with emergencies.In the case of frequent traffic congestion,it can help the traffic control department to adopt corresponding congestion evacuation methods to alleviate the congestion situation,and realize the congestion prediction based on the travel plan and the rapid response to the congestion.(2)Establish a congestion prediction model for urban expresswaysAiming at the characteristic of expressway without signal control and other factors,the running speed of the road section is predicted to judge the road conditions in the future.The model introduces the idea of travel planning to obtain the number of vehicles that will be integrated into the road network in the future.The travel plan data and spatiotemporal data are input into the RBF neural network,and the frog leaping algorithm is used to optimize the weights and parameters of the neural network,and the output road section.The average driving speed at the future time,and judge whether the road section is congested according to the speed.The experimental results show that the prediction results of the RBF neural network proposed in this paper are much more accurate than the traditional methods.(3)Constructing a model for distinguishing types of traffic jamsTraffic congestion type discrimination model can determine the types of congestion: frequent congestion,sporadic congestion.In this paper,it is demonstrated that the change of traffic flow parameters when the traffic flow system has occasional congestion can be explained by the pointy point mutation theory.Combined with the pointy point mutation theory and McMaster algorithm,the congestion type discrimination model of McMaster algorithm based on pointy point mutation theory is proposed,and the congestion type is discriminated by discussing the relationship between the congestion data and the biforked set.At the same time,the judgment results were compared with those of the traditional McMaster algorithm,and it was shown that the judgment results of the McMaster algorithm model based on the sharp point mutation theory were more accurate.The congestion prediction and congestion type discrimination mechanism based on travel planning proposed in this paper plays an important role in improving the smart traffic management system based on travel planning.This mechanism makes up for the system’s gaps in congestion prediction and processing,and improves the speed prediction of expressways.Accuracy and precise identification of congestion types and rapid response to traffic emergencies play an important role.At the same time,it provides a basis for traffic management departments to take control measures and accelerate the process of sustainable urban development.