| In urban road conditions,the change of Signal Phase and Time(SPaT)at different intersections makes the vehicles have frequent acceleration,deceleration and stopping,which not only affects the traffic efficiency,but also increases the fuel consumption and emissions,and makes environmental pollution and traffic congestion more and more serious.The development of Connected Vehicle provides conditions for solving these problems.Optimized speed in the environment of Connected Vehicle is used as reference speed for drivers to realize eco-driving,which can effectively improve fuel economy and traffic efficiency.Nowadays,most of the relevant studies focus on traditional vehicles,ignoring the comprehensive advantages of hybrid electric vehicles in energy saving and endurance mileage and the potential of energy management.Moreover,the adaptability of different drivers to reference speed is also insufficient.In addition,under the environment of Connected Vehicle,vehicles are no longer isolated individuals in the transportation system,and vehicles in an area can be studied as a whole.Therefore,for hybrid electric vehicle queue on the signalized road,this paper studies the speed optimization considering driving style under the environment of Connected Vehicle.Firstly,hybrid electric vehicle and energy management system model are constructed.The structure of hybrid electric vehicle is determined and the working mode of hybrid electric system is divided and analyzed.The dynamic characteristics of coupling mechanism under different working modes are analyzed by lever method,and the vehicle model of hybrid electric vehicle and the optimal operating line control strategy are established by forward modeling method in Matlab/Simulink,which provides the basis for the follow-up study.Secondly,the driving characteristics of drivers with different driving styles are analyzed.Based on the driving data collected from driving simulation experiments,the driving styles of drivers are classified into three categories: radical,normal and conservative by using principal component analysis and K-means clustering.Driving characteristic analysis is carried out by driver’s following experiment on driving simulator,and driving style coefficients corresponding to different driving styles are obtained,which can be used as the basis for solving reference speed.Furthermore,a driving style recognition method based on support vector machine is proposed,which realizes the recognition of different driving styles.Thirdly,the road scene model and the fuel consumption model of hybrid electric vehicle based on experimental data are established,and the speed optimization method of hybrid electric vehicle queue is proposed,including the initial target speed solution and the final target speed solution.Initial target speed is solved based on traffic information such as traffic lights,which realizes green light pass and maximize traffic efficiency.On this basis,further considering fuel economy,driving style,safety and other factors,multi-objective optimization problem is constructed and particle swarm optimization algorithm is used to solve the final target speed to achieve the speed optimization of hybrid electric vehicle queue.Finally,in order to validate the effect of the speed optimization algorithm in this paper,simulation scenarios and parameters are designed,and simulation tests are carried out based on Matlab/Simulink environment.The speed optimization method in this paper is compared with the Gipps car following model without speed optimization in ordinary environment,and the actual acceleration of the vehicle is compared with the expected acceleration corresponding to the driving style.The results show that the speed optimization method in this paper can make the vehicles avoid the red light and reduce fuel consumption by 16.7% and travel time by 4.9% while ensuring driving safety,and can meet the adaptability of drivers with different styles. |
