Research Of Vehicle Platoon Formation Based On Clustering

With the development of science and technology,the number of motor vehicles is growing rapidly.The following problems such as traffic accidents,traffic jams,air pollution and energy waste have become important problems.In order to improve the road carrying capacity,improve the road traffic efficiency,reduce the energy consumption of vehicles and reduce the air pollution to protect environment,vehicle platoon formation has become an important driving mode in road traffic,it is also an important research topic in the academic field.In this paper,the vehicle platoon formation process based on clustering algorithm is studied.Under the framework of vehicle network based on MEC,the platoon formation optimization process with energy consumption as the target is realized.The specific research contents of this paper are as follows:(1)A network architecture of formation vehicles based on MEC is introduced,and the scene of platoon formation and the formation model of platoon vehicles are presented.Aiming at the platoon formation process,the MEC logic based on formation application is described.(2)According to the traditional K-means clustering algorithm,the platoon formation algorithm is designed which includes two parts.The design idea is as follows: Firstly,the individual vehicles are used as the data individuals,every first vehicle of each platoon is selected as the cluster center to realize the first clustering;Then the platoon queues are regarded as the data individuals,the queue is selected as the cluster center,and the merging among platoons is realized by the strategy of breaking,separating and merging.The dynamic real-time updating of platoon formation is realized without considering the specific motion control of the vehicles,and finally all platoons become one queue.(3)Based on the designed clustering algorithm and an improved bacterial foraging algorithm,the formation process of vehicles is studied.The research process is as follows: The kinematic process of the formation vehicles is regarded as an optimization process,the instantaneous energy consumption index is taken as the optimization target,and the acceleration input in the formation process of the connected vehicle is optimized by the improved bacterial foraging algorithm.Combined with the clustering algorithm,the platoon formation matrix provided by clustering algorithm is used to give different air resistance coefficient gain values for vehicles in platoon and vehicles not in platoon to form one platoon.The bacterial foraging algorithm transmits the optimized acceleration control information to the clustering algorithm.The clustering algorithm uses the acceleration information to update the vehicle status information and vehicle platoon formation matrix,and then the bacterial foraging algorithm takes this information as the input until all vehicles become one platoon queue.(4)According to the above platoon formation model,clustering design and formation optimization algorithm,the platoon formation simulation research is carried out.The clustering algorithm and bacterial foraging algorithm are combined,and the numerical simulation is carried out in MATLAB and Python environment.The results show that the formation process of vehicles can achieve the optimal energy consumption under limited conditions in the proposed scenario,which verifies the effectiveness of the vehicle platoon formation model and energy consumption optimization algorithm designed in this paper.