Research On Adaptive Resource Scheduling Algorithm For Multi-vehicle Formation In Internet Of Vehicles

With the rapid development of information technology,the application of Internet of Vehicles as the Internet of Things in the field of modern transportation has been widely concerned in many countries and regions.In recent years,with the rapid increase of the number of motor vehicles in China,it is difficult for China's existing transportation network resources to adapt to its growth rate,and to a certain extent,it will bring non-negligible safety hazards,traffic congestion problems and energy consumption problems to the people.Intelligent assisted driving With the continuous development of the Internet of Vehicles technology,real-time sharing of information between people and vehicles has become possible,increasing the traffic rate of roads and saving energy and reducing emissions.Adaptive Cruise Control(ACC)has also emerged.To a large extent,it solves a series of traffic safety problems caused by the growth of car ownership.Based on the analysis of a large number of vehicle formation,adaptive cruise control,and resource scheduling methods,this paper designs a solution to the shortcomings of the existing methods in the field of artificial intelligence: deep learning and Q learning.An adaptive cruise algorithm based on deep reinforcement learning.At the same time,the problem of computational task processing for delay perception in the fleet is studied.An adaptive resource scheduling scheme based on Lyapunov algorithm is proposed.The specific work of this paper is as follows:(1)Research on fleet cooperative adaptive cruise control algorithm based on deep reinforcement learning.Firstly,in order to overcome the slow convergence problem of the traditional reinforcement learning method in solving the ACC problem,the reconstruction of the framework problem based on reinforcement learning is carried out.Secondly,in order to solve the convergence problem in the Q learning method,the deep reinforcement learning algorithm DQN of artificial neural network is introduced..The experience return method is used to further improve the convergence speed of the algorithm.Through simulation comparison and analysis,the proposed algorithm reaches the convergence domain earlier,and the performance of the QL algorithm is improved by about 21% compared with the QL algorithm.The feasibility of the cooperative adaptive cruise control algorithm based on deep reinforcement learning is verified.And advantages.(2)Research on adaptive resource scheduling method in fleet based on Lyapunov algorithm.In order to solve the contradiction between the limited computing resources of the vehicle and its rapidly growing traffic demand,firstly,the computing resource scheduling model within the fleet is established.When the computing resources from the vehicle are insufficient,the headstock resources can be scheduled to meet the delay requirements of the computing tasks.Secondly,starting from the real-time business needs of the vehicle,according to the available computing resource tolerance of each vehicle in the fleet,to reduce the overall energy consumption and packet loss rate of the fleet,jointly optimize the local computing resources,packet loss,and the transmit power at the time of unloading.Vehicle computing resource allocation strategy.Finally,the simulation performance analysis of the proposed adaptive resource scheduling scheme based on Lyapunov algorithm is carried out,and the effectiveness of the algorithm is proved.