火力分配多目标规划模型的改进MOPSO算法  

Improved MOPSO algorithm for multi-objective programming model of weapon-target assignment

作　　者：刘晓[1] 刘忠[1] 侯文姝[2] 许江湖[1] 

机构地区：[1]海军工程大学电子工程学院 [2]中国人民解放军91919部队

出　　处：《系统工程与电子技术》2013年第2期326-330,共5页Systems Engineering and Electronics

基　　金：中国博士后科学基金(20090461460)资助课题

摘　　要：提出一种改进的多目标粒子群优化算法(multi-objective particle swarm optimization,MOPSO)算法,通过化解约束条件、修改速度和位置更新等使该算法适于求解火力分配多目标规划模型。最终求解的非劣解集构成Pareto前沿,体现增加火力单元数量对射击效能的影响,决策者可按照意图从中选取最终解。不考虑多目标规划模型中的属性目标,对敌毁伤概率随迭代步数演变与单目标函数相比,收敛性能相同,最大值相近,验证了所提算法的有效性。An improved multi objective particle swarm optimization （MOPS（）） algorithm is proposed, which adjusts constraint conditions, modifies the velocity and position update formulas for the multi-objective programming model of weapon-target assignment optimization. The algorithm can obtain a non-inferior solution to form the Pareto front, which can reflect the effect of increased firepower unit on firing efficiency. Decision- makers can find the final solution from the non inferior subset solution according to their intention. Compared with the single objective function in the iterate evolution, the maximum value of the kill probability to enemy shows the same convergence performance and close maximum result, excluding the factor obieetive of the multb objective programming model, which verifies the effectiveness of the MOPSO algorithm.

关 键 词：火力分配 多目标规划 多目标粒子群优化 反导 

分 类 号：V247[航空宇航科学与技术—飞行器设计]