Research And Application Of Vehicle Safe Lane Changing Model In Traffic Simulation

At present,there are some common problems in microscopic traffic simulation.One is that the vehicle lane-changing model used in the current micro-traffic simulation system is still relatively simple and cannot reflect the main characteristics of urban traffic flow;the other is that the vehicle lane-changing model used in the current simulation system lacks consideration of safe driving behavior,vehicle collisions are prone to occur in the process;third,the vehicle lane-changing model adopted by the current simulation system does not clearly distinguish the lane-changing rules between the road section near the intersection and the ordinary road section,resulting in a lack of simulation reality.Therefore,this thesis takes the vehicle lane-changing model as the core,researches the vehicle safe lane-changing model under mixed traffic flow,and applies it to the selfdeveloped urban micro-traffic simulation system.The main works of the thesis are listed as follows:(1)In this thesis,a vehicle safe lane changing model for a mixed traffic flow in a general lane scenario was proposed.This model addresses issues such as the simple vehicle lane-changing model in the current micro-traffic simulation system,which cannot reflect the characteristics of traffic flow,and the lack of attention to safety in the vehicle lane-changing model.The improved algorithm based on the cellular automaton model is used to consider vehicle types,factors such as internal and external road traffic capacity,driver type,safety,etc.,have achieved the effect that the simulated vehicle can safely and accurately change lanes.And through the simulation experiment,it is proved that the model can reflect the main characteristics of the road traffic flow;the model before and after the improvement is tested separately and compared with the real data in the NGSIM data set.The simulated traffic flow using the improved model is compared with the actual traffic flow.The results show that the traffic flow error is 4.28%,and the lane change error is 3.31%.Compared with the model before the improvement,the accuracy is significantly improved.(2)In this thesis,a vehicle safe lane changing model in the scene of a road section near an intersection was proposed.Aiming at the problem that the vehicle lane-changing model in the current simulation system does not clearly distinguish the lane-changing rules between ordinary road sections and adjacent intersection road sections,resulting in poor simulation accuracy,an improved algorithm based on cellular automata is used to consider vehicle coordinates,the lane setting and other factors have realized the effect of safe and accurate lane change of the simulated vehicle in the road section near the intersection.And through simulation experiments,it is proved that the model can reflect the characteristics of the traffic flow of the adjacent intersections,and the comparison experiment under the same scene with Sumo,the error of the number of lane changes is5.4%,which is more accurate than the model before the improvement has improved.(3)In this thesis,a set of urban microscopic traffic simulation prototype system was designed and implemented.The system adopts a layered design,from the bottom to the top are the traffic simulation engine,the traffic simulation business system platform,and the traffic simulation front end.As the core level of the system,the traffic simulation engine adopts the vehicle lane change model proposed in this thesis,and is responsible for providing specific traffic simulation functions.In addition,the system also provides such as custom data submission,simulation model modification,and simulation result analysis,simulation data storage,simulation engine module management,user management and other functions,and the usability of the system has been proved through tests.