Trajectoty Optimization Methord Inside Autonoumous Vehicle Zone Considering Vehicle Heterogeneity

With the rapid growth of economy,the number of urban vehicles has gradually increased in China,and the traffic congestion is becoming more and more severe.As an effective solution,with the development of automation technology,autonomous vehicles have been widely concerned by scholars,and become the trend of future traffic.At the same time,vehicle's lane changing involves both longitudinal and latitude speed changes and movements,which is a key scene in autonomous vehicles.However,there are still some problems and shortcomings in the study of autonomous vehicles in the environment of real road network.At present,there are few researches on the lane changing of the autonomous vehicles,and the existing methods cannot solve the vehicles' trajectory optimization problem of real road network with mixed traffic flow of heterogeneous autonomous vehicles.In view of this reality,this paper studies the hybrid traffic management method based on vehicle trajectory optimization inside a heterogeneous autonomous vehicle zone.Based on the theory of dynamic traffic distribution,this paper takes use of integer programming model to solve the optimal vehicle trajectory of heterogeneous autonomous vehicles.Considering both forced and free lane changing behaviors between heterogeneous autonomous vehicles in a highway section,combined with the limited speed conditions of each lane of expressway,the flow diversion of fast and slow vehicles is accomplished.On the basis of ensuring safety and efficiency,the Traffic management of autonomous vehicles in the expressway section is realized.The vehicle trajectory optimization model takes the times of lane changing and the total traveling time as objective function and discretizes the time into unit time point in 1 second.With Consideration of the vehicle motion constraints,consistency constraints among variables,Lane speed restriction,free lane changing constraints and conflict avoidance constraints,The vehicle trajectory optimization model is formulated as a mixed integer program,then solved by the off-the-shelf solvers: CPLEX,to obtain the lane,displacement,and time information of each autonomous vehicle.Finally,the trajectory of each vehicle is further graphitized,and the optimized vehicle trajectory graph is obtained to realize the traffic management of heterogeneous vehicle flow in highway.Furthermore,some numerical experiments on this model are carried out.According to the traffic information in the real road network,the parameters are set,we discusses and analyzes the undetermined coefficients in the target function,and summarizes the relationship between the target function value,the number of lane changing and the coefficient of vehicle's lane changing cost.Finally,the best vehicle lane-changing loss coefficient is obtained according to the relationships between them.Then,the vehicle trajectory optimization is solved using this model,with both safety and efficiency,and the reliability of the model is verified too.