Traffic Forecasting Of Highway Toll Station Based On Spatio-temporal Correlation Of The Networked Toll Data

Accurate mastery of the trend of traffic flow at highway toll stations is a prerequisite for effective management and timely implementation of management and control measures by the transportation department.Accurate forecasting of highway toll station traffic is conducive to traffic planners to carry out path planning,traffic managers to limit the flow rate,traffic decision makers to scientifically schedule toll station personnel and reasonably switch the charging window.Most of the existing researches on high-speed traffic are based on cross-sections.Most of the research on toll stations is based on the time correlation of toll station traffic.It is difficult to deal with the impact of emergencies on traffic flow,so as to accurately predict highway tolls.Station traffic flow.With the development of modern traffic information collection technology,the full coverage of highway network toll data,combined with the time and space relationship between highway toll stations,is analyzed and researched,and the toll station traffic is predicted to alleviate traffic congestion,achieve path selection,and charge.Reasonable allocation of station resources has theoretical and practical significance.By analyzing the time correlation of toll station traffic,this paper establishes a time-based traffic prediction model.Through the analysis of the spatio-temporal relationship of highway toll stations,a flow transfer coefficient calculation considering vehicle delay and discreteness is proposed.Based on this method,a toll station traffic prediction model combining time-space correlation and time correlation is established to predict the traffic volume of highway toll stations.The main contents include:(1)Time-correlation based toll station traffic prediction method.For the existing time-based toll-based toll reporting methods,most of them are linear methods,which are simple in structure and poor in prediction accuracy.To this end,this paper proposes a time-dependent analysis of the toll station traffic data,and proposes a LSSVM model optimized by the PSO method to forecast the traffic flow of the highway toll station.The experimental analysis of Chongqing highway expressway charging data is carried out.The results show that the proposed method has applicability and effectiveness.(2)A method for analyzing the spatio-temporal relationship of toll-to-station flow between vehicles with time delay and discreteness.Aiming at the spatial-temporal relationship analysis between existing toll stations,the discreteness and delay of the vehicle are analyzed without considering the discreteness of the vehicle and the imperfection of the vehicle delay.The discreteness and time-delay of vehicles are analyzed,and the analysis method of space-time relationship between highway toll stations considering vehicle time-delay and discreteness is put forward.The experimental verification of the historical highway toll data in Chongqing is carried out.The experimental results verify the effectiveness of the proposed method.(3)Toll station traffic combination forecasting method combining time-space correlation and time correlation.Based on the previous estimation results,a traffic prediction model of highway toll station based on the combination of time correlation and space-time correlation is established.Finally,the Chongqing highway network charging data is selected to analyze the method proposed in this paper,and its applicability and effectiveness are verified.In summary,the highway toll station traffic prediction model based on the spatio-temporal correlation of the networked toll data can reflect the impact of the sudden situation on traffic flow,and has a good prediction effect.It provides an effective reference for traffic law enforcement departments to command and dispatch,rational allocation of toll station resources and choice of travelers' paths.