Study On Lane-changing Model For Expressway Weaving Area In Microscopic Traffic Simulation

The expressway weaving area as the confluence area between main vehicles andmerging vehicles, there are a lot of lane-changing behaviors, since the merging vehiclesmust complete lane-change behaviors within weaving area, the lane-changingphenomenon is more prominent at weaving area than the ordinary section. Frequently lanechanges can easily lead to traffic Flow disorder and a significant reduction of capacity,and the improper lane-changing behavior is the main cause of traffic accidents. Due to thecomplex lane-changing behaviors at weaving area, the model is more difficult to obtain itsfeatures, which greatly increases the difficulty of modeling and affect the accuracy of thesimulation. In addition, lane-changing model as the one of the most importantmicroscopic traffic simulation model for describing driving behavior, due to thecomplexity of lane-changing behavior and the microscopic data of vehicle operation isdifficult to obtain, the study on lane-changing model is still very weak in our Country.Previous studies in related fields at home and abroad have made certainachievements, but fewer and scattered on the whole, the shortcomings of existing relevantresearch works is in the following three areas: firstly, current researches on thecharacteristics of the driver focused primarily on traffic safety macroeconomic statisticalanalysis, while ignored the ambiguity by type of driving behavior, and without associatewith actual driving behavior, the classified results cannot be applied in microscopic trafficflow simulation; second, most of the existing simulation models assume that the mergingvehicle does not affect the mainline vehicle, or merging vehicle will be “humility”mainline vehicles, so that the complex dynamic interactions between the two traffic floware simplified, and which leads to the simulation results not agree with the actual situation;third, as the lane-changing model parameter calibration requires a lot of microscopictrajectory data, the data is very difficult to obtain, the most of lane-changing models ofmerging area are stay in the modeling ideas and conceptual framework, so that thelane-changing model cannot be used in the simulation model, and the applicability of themodel is scant. Therefore, this paper summarizes the advantages and disadvantages of the existingresearch of domestic and foreign, and the survey and experimental research, systemsanalysis, computer simulation and theoretical analysis, logical reasoning and other relatedmethods are used. Based on the survey and analysis of traffic flow and driverscharacteristics at expressway weaving area, a continuous driver type integration model isbuilt. The study work contents were showed as follow.(1) Nine typical expressway weaving areas are selected to survey, the videoprocessing software VEVID is used to deal with the investigation Videos, and a lot ofvehicle trajectory data is got, then to analyze the traffic flow and lane-changingcharacteristics by use of the principle of mathematical statistics.(2) The driver characteristics are analyzed by using driver behavior scale surveymethods, and get the influence of the driver characteristics on driving behavior, and thepotential behavior related factors; then combines with the uncertainty and ambiguity ofdriver types classify, the Fuzzy clustering theory is introduced, and a continuous drivertype integration model is built by the fuzzy clustering method and K-S test method.(3) The interaction between traffic flows characteristics are analyzed through thevehicle trajectory data. Then the courtesy and cooperation mechanism are introduced, thecooperative lane-changing model, forced lane-changing model and gap acceptance modelare built. Based on the survey traffic data, the factor analysis and stepwise regressionmethod are used to select the model input variables, and the model parameters arecalibrated by use of maximum likelihood estimation method.(4) The microscopic traffic simulation system TPSS is taken as simulation platform,and combination of software engineering design and C++Programming Language, thesimulation module of lane-changing model of weaving area is built, and the RMSE andMPE evaluation indexes are choose to validate the built model. The validation results areindicate that the built model can be used to describe the impact of driver type on drivingbehavior.(5) Considering the mainline and ramp service traffic flow, the effect of differentlengths of the acceleration lane on traffic flow service levels is analyzed by use of builtsimulation model, and the recommended values of the acceleration lane under a differentlevel of service are obtained.The main innovation of this research work is shown in the following three aspects: (1) The effect of the driver characteristics on the traffic flow behavior is quantified,and combines with the uncertainty and ambiguity of driver types classify, the Fuzzyclustering theory is introduced, and a continuous driver type integration model is built bythe fuzzy clustering method and K-S test method, which can represent the effect of thedriver characteristics on the traffic flow behavior really.(2) The courtesy and cooperation mechanism are introduced, and the effect of thedriver characteristics on the traffic flow behavior is considered, the cooperativelane-changing model, forced lane-changing model and gap acceptance model are built.(3) According to a large number microscopic operating data, the factor analysismethod and stepwise regression method are used to select the variables which reflectlane-changing information and impact significantly as the input variables of models, andthe models parameters are calibrated by use of maximum likelihood estimation method,overcome the deficiencies of empirical calibration method.The simulation model which can accurately represent the traffic flow characteristicsof expressway weaving area is built through this study work, and can provide technicalsupport to alleviate urban expressway congestion problem. The simulation model has animportant theoretical and practical value to promote the application and implementationof intelligent traffic technologies in the transport sector.