Research On Collaborative Air Combat Strategy Of Unmanned Combat Aerial Vehicles

With the more and more universal application of UCAVs in the war, UCAVs coordinated aircombat decision has become a hot research topic. Making a good battle plan in the shortest timewill provide an effective protection for the success of UCAVs air combat. According to the process,the battle can be divided into beyond and within visual range air combat. In this paper, groupingdecisions are researched.Coordinated attacking tactic of our UCAVs during beyond visual rangeperiod and coordinated maneuvering strategy during within visual range period are studied. Foreach question, corresponding model is set up and different optimization algorithms are designed tosolve the models. In order to prove the pros and cons of the algorithms, several examples aresimulated in software Matlab and MAK to display the air combat effects.For the UCAVs grouping part, the enemy’s grouping model is set up based on the principlethat the UCAVs with similar attacking characteristics are classified into one group, and improvedfuzzy c-means clustering algorithm is proposed to solve this model. Compare the improved andoriginal fuzzy c-means clustering algorithms. Our UCAVs grouping model is built based on theprinciple of maximizing winning margin. Combine genetic algorithm with simulated annealingalgorithm to work out the model. Compare this algorithm with genetic algorithm and simulatedannealing algorithm.The results show that improved fuzzy c-means algorithm can figure out bettergrouping strategy for enemy UCAVs and genetic simulated annealing algorithm can seek bettergrouping strategy for our UCAVs.For coordinated attacking strategy during beyond visual range period, the situation threats areestimated and weights of each threat are analyzed by analytical hierarchy process and grayassociation method. Based on the principle of minimizing the expected remaining threat,coordinated attacking model is established and improved genetic simulated annealing algorithm isproposed to figure out the model. Apply improved and original genetic simulated annealingalgorithm to examples. The results show that improved genetic simulated annealing algorithm cansearch better attacking strategy.For] coordinated maneuvering strategy during within visual range period, the UCAVmaneuver model is built. Prediction influence diagram is applied to simulate the process of twoUCAVs maneuver against each other, and group decision theory is used to figure out UCAVsmaneuver decisions. Simulation results show that prediction influence diagram and group decisiontheory can effectively guide UCAVs to fight with each other.