Mining And Modeling Of Lane-changing Behavior Based On Trajectory Data

Lane-changing is one of the basic driving behaviors which may induce traffic oscillations and incidents.However,it is difficult to well model the lane-changing decision process due to the complex traffic conditions on the actual roads.Similarly,the development of the lane-changing decision model is also slow.The existing models start from the driver's lane-changing probability and build a probabilistic lane-changing decision model based on the utility theory.To promote the prediction accuracy of lane-changing decisions,this paper present probability lane-changing decision model by taking into account the memory effect and driver heterogeneity,respectively.That is,the former lane-changing decision model considers a series of trajectory data rather than the data of a specific time utilized in most existing models.The drivers are classified in terms of lane-changing trajectories in the latter which is expected to construct a multi-type lane-changing decision model.Furthermore,a probabilistic lane-changing decision model comprehensively considers the memory effect of the lane-changing trajectory data and driver heterogeneity which is expected to further promote the prediction accuracy of the lane-changing decision model.Calibrations and validations are carried out based on the NGSIM data,which indicate that the proposed model can significantly promote the prediction accuracy of lane-changing decisions.In addition,a probabilistic lane-changing decision model that considers memory effects and driver heterogeneity further promote the prediction accuracy of the lane-changing decision model.The main research contents of the article are as follows:(1)To select critical factors of lane-changing decisions,the binary logistic regression analysis is carried out.The independent variables,the important 13-factors influencing the lane-changing decisions,are selected,including the acceleration of the lane-changing vehicle,the space headway and the speed difference between vehicle n and the immediate preceding vehicle in the current lane,and the space headway and time headway between the vehicle nand the immediate preceding vehicle in the target lane and the immediate following vehicle in the target lane.Extracting lane-changing related data from NGSIM(Next Generation Simulation)data for preprocessing.(2)Based on the probabilistic lane-changing decision model,a probabilistic lanechanging decision model considering the memory effect of the lane-changing trajectory data is constructed.This model significantly improved the prediction accuracy by about 20% due to taking advantages of more historical data.The results show that the probabilistic lane-changing decision model considering the memory effect can significantly improve the prediction accuracy of the lane-changing model.(3)Based on the probabilistic lane-changing decision model,a probabilistic lanechanging decision model considering the driver's heterogeneity was constructed.This model can improve the prediction accuracy by about 15% compared with the traditional model,indicating the advantages by considering driver heterogeneity.(4)Based on the probabilistic lane-changing decision model,a probabilistic lanechanging decision model that considers lane-changing trajectory data and driver heterogeneity was constructed,which further improved the predictability of lanechanging.The prediction probability was 91.83%.The results show that the memory effect is considered The probabilistic lane-changing decision model with driver heterogeneity can further improve the prediction accuracy of the lane-changing model.