| With the continuous improvement of the performance of air strike weapons and the continuous innovation of air strike combat modes,the traditional “platform-centric” air defense combat system is unsustainable,and future wars are bound to shift from “platformcentric warfare” to “network-centric warfare”.As a key part of the air defense mission performed by the warship formation,Weapon target assignment(WTA)problem is currently solved by centralized computing algorithms.Under the background of “networkcentric warfare”,centralized computing algorithm faces problems such as high communication pressure,large-scale problem and single-point failure.In order to solve the above problems,This thesis studies the distributed computing framework for weapon target assignment to solve the multi-platform weapon target assignment problem,and proposes the model solving algorithm based on distributed optimization and the model solving algorithm based on distributed auction mechanism.Combined with the model solving algorithm based on distributed optimization,an auxiliary decision-making system is designed to support dynamic weapon target assignment.The model solving algorithm based on distributed optimization aims to reduce the threat of the incoming targets and the resource consumption of the warship formation.Each combat platform submits multiple interception schemes to the central node,and then the central node selects parts from the local interception scheme to form a global interception scheme.In this thesis,the NSGA-II algorithm is used to solve the multi-objective optimization problem of the combat platform and the center node.When the platform performs optimized calculations,Logistic mapping is used to ensure the diversity of the initial population? the archiving mechanism is used to save the outstanding individuals appears in the local optimization process,the mechanism for local optimization to stop running is designed according to the proportion of non-dominated solution sets in the archive? the selection strategy of local interception schemes is designed by referring to the concept of knee points in multi-objective optimization.Experimental results demonstrate that the model solving algorithm based on distributed optimization has certain advantages in terms of running time and can guarantee the quality of the interception scheme to a certain extent.The model solving algorithm based on distributed auction mechanism aims at maximizing the interception probability.In each iteration,each firepower unit constructs a local target set based on a greedy strategy,and eliminates each firepower unit’s differences in the interception schemes by exchanging information with adjacent firepower units.After multiple rounds of iterations,multiple firepower units converge to the same interception scheme.To support a compound strike scenario where a maximum of two firepower units intercept the same target,the shadow targets are introduced into the algorithm,and a strict data update strategy is designed for it.Experimental results verify the convergence of the model solving algorithm based on distributed auction mechanism and the quality of the interception scheme.Considering the actual air attack scenarios with multi-batch and multi-target,an auxiliary decision-making system supporting dynamic weapon target assignment is designed and implemented on the ground of the model solving algorithm based on distributed optimization.The system constructs the target queue based on the sequence of the launch window of the incoming target,and realizes the dynamic weapon target assignment in stages based on the target queue. |
