Research On The Region Air-Defense Campaign Command Decision Based On Swarm Intelligence

Some researches on the theory and applications of the Region Air-Defense CombatCommand Decision problems, which focus on the forces deployment and weapon-targetassignment, are done.Firstly, the characteristics of tactics, techonology, and campaign employment of typicalattacking weapons such as tactical ballistic missile(TBM), cruise missile(CM) andair-to-ground missile(AGM) are analized here, and also the disadvantage of these two missilesin modern air attack combats. Basic principles of the region air-defense combat forcesdeployment is summarized. The models of the region air-defense combat forces deploymentbased on the principles of maximizing air-defense effectiveness and minimizing costs areproposed.After analizing the weapon-target assignment problem of the region air-defense combat oncondition that the air attack is big-scale, we proposed the models of weapon-target assignmentbased on multi-objects evaluating function, considering the probability of successfully defendingthe region, maximizing the number of destroyed targets, and minimizing the costs.Secondly, the relative theories of swarm intelligence optimization are analized. To solvingthe weapon-target assignment problem more efficiently, a new hybrid algorithm based on tabusearch and particle swarm optimization is proposed. The efficiency of the new algorithmis proved by simulations of non-uniform scale of the problem.In order to accurately solve the forces deployment problem of the region air-defense in largeair-attack scale, an accurate multi-agent tabu particle swarm optimization algorithm is proposed,introducing the competition, cooperation and self-study mechanisms of multi-agent technology.The effectiveness and efficiency of the new algorithm is also proved by simulations in differentproblem scale.At last, the theories and implements of the region air-defense combat command intelligentdecision support system (RADCCIDSS) are studied. The contents, work flows and the wholefunction of decision module are researched, so do the database, model base, knowledge baseand theirs management systems of the RADCCIDSS. The development method of theRADCCIDSS is discussed, while the RADCCIDSS is implemented partly.