Traffic Efficiency Analysis And Optimal Control Method Of Urban Expressway Weaving Area

With the rapid development of urban expressway construction,there are some problems in the actual operation process,which leads to the traffic efficiency not reaching the expected level.Due to the long-term and large amount of actual data is difficult to obtain,this paper uses the method of combining actual data and simulation to explore the traffic operation law of urban expressway weaving area,analyzes the factors affecting the traffic efficiency of weaving area,and applies the variable speed limit guidance control method based on DQN algorithm,which can effectively improve the traffic efficiency of urban expressway weaving area.Firstly,this paper selects the micro traffic simulation software VISSIM to establish the urban expressway interweave area model,uses the actual survey data,and designs the model parameter verification scheme by using the Latin hypercube sampling(LHS),and verifies the model effectively,which makes the weaving area model and the actual weaving area operation condition infinite close.Taking the interleaved area as the research object,the paper analyzes the traffic efficiency and the influencing factors of the weaving area under the different traffic flow input of the main line,which provides the theoretical basis for improving the traffic efficiency of the weaving area of the expressway.Secondly,considering that the average vehicle travel time per unit length in the weaving area is an important embodiment of the traffic efficiency and service level of urban expressway weaving area,the influencing factors of the traffic efficiency in the weaving area(main line input flow,relative Import ramp ratio,exit ramp discharge ratio,weaving area length)and the average travel time of the unit length of the vehicle in the weaving area are established by SPSS analysis In addition,the traffic operation state corresponding to the steep drop of the capacity of the weaving area is taken as the traffic efficiency under the expected level,and the length of the weaving area is obtained based on the expected level of the traffic efficiency,based on the reasonable prediction of the upstream flow and the relative inflow ratio of the inlet ramp in order to ensure the traffic efficiency of the weaving area The recommended value of degree,from the design point of view,puts forward the measures to improve the efficiency of weaving area effectively.Finally,based on reinforcement learning,the traditional Q-learning algorithm is improved,and the variable speed limit guidance control method based on DQN algorithm is applied.The variable speed limit value is taken as the action set,the variable speed limit area density and interweaving area density are taken as the state space,and the sum of average delay and parking times of vehicles in road network is taken as the return function.By constructing VISSIM Python simulation platform,taking the North Ring Interchange Section of Chongqing inner ring expressway as the simulation background,and finally using the road network performance indicators(total travel time of vehicles,average delay of vehicles,average travel time of vehicles in the weaving area,traffic volume passing through the weaving area)and vehicle trajectories for evaluation and analysis,it is proved that the variable speed limit guidance control method based on DQN algorithm can be effective it can improve the traffic efficiency of weaving area.