Distributed Multi-type Uav Task Allocation And Simulation

With the increasingly complex battlefield environment,the missions and the scope of the UCAV grow variously.The operational manner of UAV changes from the independence towards the multi-way,more types of UCAV.Compared to the single campaign mode of the UAV,the operational tasks must be finished coordinately by a number of vehicles,different types of UAV in multi-type UCAVs.Through the research of the campaign system and task assignation of the multi-type UAVs,you can make the types of completed tasks more diversely and the quality and efficiency improved greatly.In order to utilize and coordinate the resource of multi-type UCAVs better,and improve campaign efficiency of multi-type UCAVs,the problem of targets clustering,task assignation was studied deeply.In the background of SEAD and the attack as a priority,allocated the different objectives to the different UCAVs.The problem of task assignation is equal to objective assignation.The contributions of this thesis are as follows:(1)in accordance with the scope of operations and of the characteristics of a more decentralized objectives,it research of the target clustering method based on the geographical location.The insufficient of traditional K-means algorithm in solving the problem and the improved K-means clustering algorithm are analysed.The simulation Of traditional methods and improved methods is carried out and the results of the simulation was quantitative evaluated.It shows that the improved K-means is better than the traditional K-means.(2)in pre-distribution phase,through multi-objective optimization theory,the task assignation is described to multi-objective optimization problem.The multi-objective optimization method of leading the commander interest information and keeping diversity are proposed.The simulation of this algorithm shows that the algorithm PIMOEA can effectively address the issue of task assignation.The commander interest information reduces the burden of commander and it was a practical application of the method.(3)A group is studied in re-distribution phase.System framework is given in a group. Multi-Agent based distributed collaborative auction task assignation algorithm,which is a distributed and for the purpose of "real time" dynamic allocation algorithm.Pre-set is used before the use of algorithm and the feasibility of the algorithm is proved by simulation.