Research On The Simulation Model Of Freeway Weaving Segment Vehicle Lane-changing Based On Cellular Automata

The freeway weaving segment is the core component of the freeway for which is important to ease the congestion of freeway segment urban traffic network.The traffic condition of freeway weaving segment is complex,the traffic flow is often weaving,and the traffic congestion is easily caused by the frequent lane-changing of vehicles,affecting the smooth moving of vehicles in weaving segment.Therefore,analyzing its influence on the running state of the vehicles of weaving segment through the study of lane-changing characteristics,and establishing weaving segment lane-changing prediction model of traffic flow condition and constituting traffic management plan is of theoretical and practical value to alleviate traffic congestion and improve traffic flow state in weaving segment.In this paper,the freeway weaving experimental segment,which is north of Huayuan Bridge of the West Third Ring freeway in Beijing,were selected by artificial video measurement method,and three kinds of 2 hour continuous observation data are selected,which are corresponding to free flow,the critical state of traffic flow and the traffic congestion state.Then,the data are compared and analyzed within the characteristics of traffic flow and lanechanging behavior under different traffic flow states.Based on the idea of cellular automata modeling,combined with the actual traffic operation characteristics,the weaving segment lanechanging simulation model is constructed.The model includes cell module,vehicle module,weaving segment module and data statistics module.The model parameters are selected under the model evolution rules,which including the basic parameters of weaving segment,vehicle length parameters,vehicle speed parameters,maximum speed parameters,the probability of vehicle entering the cell,the probability of lane changing,the probability of changing to the left lane,and the probability of the random slowing.To make up for the deficiency of previous research on the evolution rules of cellular automata,which has adjusted the reduction step,different path lane changing rules and lane changing conditions,location update principle,lane change probability steps according to the actual situation of different traffic flow characteristics.Finally,based on the principle of updating the two-dimensional array of vehicle information,the weaving segment lane-changing cellular simulation model system(WSCSIM V1.0)is implemented with MATLAB.The simulation platform has the visual operation interface,and the parameter is easy to input and debug.According to different traffic flow state,several experiments with calibration of parameter combination are carried out to simulate the vehicle performance.The average error between the simulation data and the actual data is not more than 10%,and the vehicle operation display trail show that the simulation results are in good agreement with the actual situation.