改进分解进化算法求解动态火力分配多目标优化模型  

Improved Decomposition-Based Evolutionary Algorithm for Multi-objective Optimization Model of Dynamic Weapon-target Assignment

作　　者：张滢[1] 杨任农[1] 左家亮[1] 景小宁[1] 何贵波[1] 

机构地区：[1]空军工程大学航空航天工程学院,陕西西安710038

出　　处：《兵工学报》2015年第8期1533-1540,共8页Acta Armamentarii

摘　　要：战前制定合理的火力分配(WTA)方案,可以优化资源配置,用最小的代价获取最大的战场收益。其一,建立了面向多型武器协同进攻作战的动态火力分配(DWTA)多目标优化模型,由多个阶段静态模型构成,各阶段静态模型参数需根据战场态势实时获取;其二,重点研究阶段静态模型求解算法。针对模型特点,设计了一种满足资源约束的编码方式,融合禁忌搜索和拥挤距离策略,提出了一种改进分解进化算法。对比实验验证了算法的可行性、快速性和有效性。A reasonable weapon-target assignment（WTA） scheme is developed to optimize the allocation of limited resources, which brings the maximum awards with minimum costs. A dynamic weapon-target assignment（DWTA） multi-objective optimization model is established, especially for the offensive opera- tion with muhi-weapon. The proposed model is composed of several stage static models. The parameters of each stage model are obtained from the battlefield in real time. An algorithm is elaborately studied to solve the stage model based on the hypothesis to deal with the dynamic model. According to the charac- teristic of the proposed model, an encoding mode is constructed to satisfy the resource constraints. An im- proved decomposition-based evolutionary algorithm is proposed by mixing of tabu search and crowding distance strategies. Comparative experiments prove that the proposed algorithm is feasible, fast and efficient.

关 键 词：兵器科学与技术 多目标优化 动态火力分配 分解进化算法 禁忌搜索 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]