Research And Implementation Of Message Transmission Mechanism In Electric Vehicle Charging And Swapping Network

With the increasing number of electric vehicles(EVs),the demand for massive EVs charging is adding to the load on the grid.In order to enable power operators to quickly and accurately btain the power demand of the EVs for charging services and improve the quality of charging services,it is necessary to design a real-time and effective message transmission mechanism between the power grid and EVs.However,the mobility and wide distribution of EVs bring a great challenge to the transmission of information between the supply and demand sides in EVCS network.The Vehicular Ad Hoc Network(VANET)can solve this problem to a certain extent by virtue of its flexible wireless networking capabilities.However,there are two problems in the application of VANET technology to message transmission in EVCS network:(1)incompatibility of communication protocols;(2)mobility of EVs leads to the instability of message transmission.Aiming at the above problems,based on the in-depth study of WAVE(Wireless Access in Vechicular Environment)protocol and routing mechanism of VANET,this paper proposes an effective message transmission scheme in the EVCS network.The main work of this paper is as follows:Firstly,this paper designs a protocol conversion mechanism,which consists of three parts: the WSMP transceiver module,the PSID(Provider Service Identifier)cache module and the protocol conversion module.To solve the problem of incompatibility between WAVE protocol and TCP/IP protocol,a p4-based programmable V2G(Vehicle-to-Grid)gateway is introduced into EVCS network,in which there are three message processing mechanisms wsmp-to-ipv4,wsmp-in-ipv4 and ipv6-in-ipv4 based on P4 are designed.Through the effective encapsulation of WAVE data packets,to achieve its transmission under the IPv4 network.Secondly,a routing mechanism based on reinforcement learning is designed to solve the problem of unstable message transmission in the mobile scenario of EVs.First,this mechanism fully considers the relationship between speed,distance,channel state and vehicle wireless link state,and designs a fuzzy logic system to make a more reasonable evaluation of the wireless link.Second,this mechanism dynamically adjusts the selection of relay nodes for reinforcement learning routing based on the results of link state evaluation,which improves its performance in routing decisions.At the same time,in order to further optimize the routing mechanism,this paper uses the K-Means algorithm to cluster the electric vehicles.By learning the route on the cluster-head node in a distributed way,the number of states in reinforcement learning can be effectively reduced and the convergence efficiency of the routing mechanism can be improved.Finally,the function of the protocol conversion mechanism is verified by building an experimental platform,and the verification results show that the functions of WSMP transceiver module,PSID cache module and protocol conversion module all achieve the expected goals.Then,the performance of the routing mechanism based on reinforcement learning is tested in the simulation environment.The performance advantage of the proposed routing mechanism in message transmission is verified by comparing with the existing schemes in three aspects: convergence speed,average end-to-end delay and packet delivery rate.