Research On Mission Planning & Attack Decision-making Under Incomplete Information For Multi-UAVs

In recent years,the UAV(Unmanned Aerial Vehicle)has become more and more essential in the military field.It plays an irreplaceable role in the war and has attracted the widespread attention of all the major military countries in the world.The operational mode of the UAV has shifted from single-UAV mode to multi-UAVs cooperative mode when faced with the diversity and complexity of the mission in the modern battlefield.In this thsis,the problem of multi-UAVs mission planning and attack decision-making are studied under the background of multi-UAVs cooperative operation.First of all,the cooperative mission planning of UAVs is analyzed.The mission area property,the tasks and the task executors are described in this part,the models of the task constraints,the map constraints and the voyage estimation are also constructed.According to the problem of multi-UCAVs cooperative attack decision-making,five threat factors: angle,distance,velocity,altitude and air combat capability are established to describe the situation of air combat and the interval number is used to represent the uncertain threat factors.Secondly,in order to solve the task assignment problem in cooperative mission planning,a kind of genetic algorithm based on the static linked list is proposed,and this method can solve the task assignment problem with multiple constraints in the mission area.The simulation results prove the effectiveness of the method.Then,A* algorithm is studied to achieve a path planning result fast and effectively in cooperative mission planning.A variety of heuristic functions of A* algorithm are analyzed and a suitable one is selected.In this part,a new storage method of the mission area map is designed and it accelerates the speed of the algorithm in the simulation.Next,because of inadequate reconnaissance in advance,temporary task adjustment and equipment failure,the task reassignment should be taken into account.The task reassignment will be triggered during the execution of the tasks when faced with: task point add,task point cancel,newly added threat areas,UAV failure and man-made task adjustment.The task reassignment method is designed using the contract net protocol,which the bidding process is improved in this part.The simulation results show that the design is reasonable.And then,considering the incompleteness of the battlefield situation information and the intention of the target,a decision-making method of UCAV(Unmanned Combat Aerial Vehicle)collaborative attack is proposed.In this part,the rough set theory is used to mine the historical air combat information for rules,and then the situation information that need to recognize and the rules are compared to get the intention.After obtaining the intention,the model of the fire distribution is constructed according to the current situation,and the Hungarian algorithm is improved so that it can be used in the current model with uncertain situation information.The simulation results show the effectiveness of the improvement.Finally,the task assignment,path planning and task reassignment are combined into a software.On the Visual Studio 2015,using MFC,a cooperative mission planning system simulation platform is developed,and the simulation process of the cooperative mission planning can be demonstrated on it.