基于记忆辅助动态UMDA的UCAV编队动态目标分配

针对无人作战飞机编队对地攻击过程中的动态目标分配问题，建立了针对联合目标的武器目标分配（weapon target assignment, WTA）模型，反映攻击方攻击意图及对目标内在关系的理解。提出基于记忆辅助的动态单变量分布估计算法（memory enhanced dynamic univariate marginal distribution algorithm, MDUMDA）对问题进行动态寻优，利用概率模型对动态寻优过程中的历史信息加以记忆和利用。仿真结果表明，所建立的WTA模型是合理的，MDUMDA能够有效求解动态WTA问题，其性能明显优于随机迁移算法。

Dynamic target assignment method for UCAV fleet using memory enhanced dynamic UMDA

Aiming at the dynamic target assignment problem in anti-ground mission of an unmanned combat aerial vehicle (UCAV) fleet, a weapon target assignment (WTA) model for associated targets is built up. A memory enhanced dynamic univariate marginal distribution algorithm (MDUMDA) is proposed to solve the problem at hand. In this algorithm, the probability model is utilized to save and reuse the historic evolutionary information. Simulation results show that the WTA model is reasonable, the MDUMDA can effectively solve dynamic WTA, and its performance is significantly better than the random immigration algorithm.