Control Methodology Of Reducing Congestion At The Freeway Lane Drop Section Under The Internet Of Vehicle

With the characteristics of high speed,safety and comfort,the freeway has been playing an important role in the modern comprehensive transportation system.Only if the freeway is not congested,its social value can come true and the inter-regional connection is strengthened.The strengthened inter-regional connection increases the freeway traffic demand,and sometimes freeway congestion occurs.Freeway congestion often occurs at bottlenecks and will propagate upstream,which can lead to the breakdown of the whole freeway or the network nearby and can lead to low traffic efficiency,low traffic safety and severe traffic polluction.Therefore,it is very important to reduce the congestion and increase traffic efficiency at freeway bottlenecks.The development of connected vehicles provides some new ideas for freeway management and control.This paper took the freeway lane drop section due to road design,road maintenance,traffic incidents and so on as the research object,analysed the congestion problem,and proposed control methodology to reduce congestion.This paper analysed the freeway traffic operation characteristics under the internet of vehicle.This paper analysed the environment of the internet of vehicle and the control measurments.The difference and relationship between connected vehicles and autonomous vehicles was analysed.Besides,this paper analysed the freeway capacity under the connected vehicle with manual vehicles,automous vehicles and mixed vehicles.Then,the freeway lane drop section capacity drop under the congested traffic state was determined under the internet of vehicle.The above analysis is the basis for the freeway short-term traffic state estimation.This paper proposed the freeway short-term traffic state estimation method.This paper established the traffic state Kalman filter model,evaluated the model using the simulated data.As the freeway traffic state changes with time varying,it can be regraded as time series data and the traffic state time series model was established.The specific traffic state GARCH model was established using the real data of Su Jia Hang Freeway.Since both Kalman filter model and time series model can estimate traffic state effectively,this paper proposed the combined method based on both Kalman filter model and time series model,and demonstrated it using real data of Su Jia Hang Freeway.The proposed method is the basis for the analysis of congestion mechanism and the corresponding control measures under the internet of vehicle.This paper analysed the traffic congestion mechanism at the freeway lane drop section under the internet of vehicle and proposed the strategy to reduce the congestion.This paper analysed the traffic congestion mechanism,including traffic congestion generation mechanism,capacity drop mechanism,congestion propagation mechanism and congestion dissipation mechanism.A numerical example was used to demonstrate the traffic congestion generation,propagation and dissipation.Besides,this paper proposed the strategy to reduce freeway congestion under the internet of vehicle,including collecting information,dealing with information and sending information back to vehicles.This paper established the variable speed limit(VSL)model at the freeway lane drop section under the internet of vehicle.This paper established the microscopic model of VSL using the traffic simulation software AIMSUN.In order to make the VSL to be optimal control,this paper determined the objective function and constraints,and proposed the improved Genetic Algorithm(GA)to solve it.After the calibration and validation of the microscopic model using the real data,this paper analysed the VSL effects under different traffic demand.Results indicated that VSL could improve the traffic efficiency at the freeway lane drop section.In order to analyse the VSL effect from different aspects,this paper established the macroscopic model of VSL and it verified the results from microscopic model.Besides,this paper illustrated the settings about VSL under the internet of vehicle.This paper proposed the lane-changing advisory control method at the freeway lane drop section under the internet of vehicle.This paper proposed the lane-changing advisory strategy under the internet of vehicle.In order to make it to be optimal control,this paper determined the objective function and constraints,and solved it using learning based Hyper-Heuristic methodology.The corresponding model was established and the control effects under differnet demand were anslysed.Results indicated that the lane-changing advisory control can reduce the total travel time and increase traffic flow.Besides,this paper anslysed the effecs with different proportions of connected vehicles.Results indicated that if the ratio is less than 10%,there is almost no effect;if it is between 20% and 40%,the traffic efficiency can increase rapidly;if it is more than 50%,traffic efficiency increases slowly.