Design And Implementation Of UAV Swarm Combat And Comunication Simulation Platform

Unmanned aerial vehicle(UAV)has the characteristics of diverse functions and flexible configuration,and has great potential and demand in both military and civilian fields.However,the performance of a single UAV is limited,and it is prone to lack of energy and external interference when performing difficult and complex tasks.Therefore,it is necessary to promote the research on the flight characteristics and control algorithms of UAV swarms.At present,most of the algorithms for swarm operations of UAVs are directly applied from the field of single-agent algorithms,and cannot be directly put into practical use.Although many institutions at home and abroad are currently doing research on UAVs,the research directions are relatively Single,either only focuses on the UAV command algorithm,or only studies the network communication of the UAV.To truly exert the power of the UAV swarm combat,it is necessary to study the impact of both aspects on the combat performance at the same time.A UAV swarm simulation platform to organically combine the UAV’s swarm combat algorithm and the UAV’s swarm communication.The emergence of the above problems has brought severe challenges to the research and practical application of UAV swarm operations.In order to solve these problems,this thesis innovatively designed a UAV swarm simulation platform,which mainly includes the following work:1.In order to solve the problem that most of the UAV swarm combat algorithms are directly applied in the field of single-agent algorithms,this paper trains and encapsulates the UAV swarm combat algorithm based on the current best multi-agent algorithm MADDPG in the MaCa environment.After completion,improve the MADDPG algorithm for the stability of the winning rate,and conduct confrontation simulation with the default multi-agent algorithm in the MaCa environment to achieve a more stable increase in the winning rate during the training process,and the upper limit of the winning rate after the training is completed..2.In order to study the combination of UAV swarm combat algorithm and UAV networking communication,aiming at the functions of fixedfrequency interference strike and frequency-hopping communication for UAV swarm communication frequency points in MaCa environment to avoid fixed-frequency strike,the The frequency hopping module,neighbor module,time slot frame module,media access layer module,6top module and scheduling algorithm module are designed based on the 6TISCH protocol under the INET framework of the OMNET++ platform,and the frequency hopping communication function of the UAV swarm is realized on the OMNET++ platform.3.In order to realize the coupled simulation of the UAV swarm combat algorithm and the UAV communication network,the MySQL database is used as a connection bridge.Every step of the MaCa environment,the location information and survival information of the UAV are written into the database.The OMNET++ platform builds a scene corresponding to the MaCa environment,and also connects with the database,reads the information of the drones in the database synchronously,designs and realizes the mobility of the drone swarm nodes in the OMNET++ platform,and realizes the MaCa environment and the OMNET++ platform.The coupled simulation of the UAV swarm is analyzed,and the simulation results are analyzed to analyze the influence of different degrees of UAV swarm command algorithm on the UAV swarm communication.The research results show that the higher the combat rate of the swarm combat algorithm,the better the communication efficiency of the UAV swarm network.