潜艇声诱饵防御声自导鱼雷改进PSO算法  

Improved PSO Algorithm to Defend against Acoustic Homing TorpedoesUsing an Acoustic Decoy of a Submarine

作　　者：侯文姝[1] 陆铭华[1] HOU Wenshu;LU Minghua(Naval Submarine Academy,Qingdao 266199,China)

机构地区：[1]海军潜艇学院,山东青岛266199

出　　处：《水下无人系统学报》2023年第3期436-441,共6页Journal of Unmanned Undersea Systems

摘　　要：潜艇使用单个小口径自航式声诱饵防御正在进行蛇形搜索的声自导鱼雷时,应快速得出防御方案使鱼雷与潜艇距离最大化。通过分析种群粒子数、迭代次数、粒子速度的上限、加速度因子和种群初始化方法等因素对基于并行计算的粒子群优化(PSO)算法的影响,确定该算法的改进方向。改进的PSO算法通过扩大粒子速度的上限以及在迭代过程中重新生成粒子群,适应度值大于7 500 m的仿真次数比改进前提高了95%,且在不增加计算量的情况下收敛得更快,达到提升解算效率的目的。A submarine that uses a single self-propelled acoustic decoy to defend against an S-type maneuver acoustic homingtorpedo must immediately implement a defensive counterplan that achieves the maximum distance between the torpedo anditself.In this study,the effects of the number of particles in a swarm,number of iterations,upper limit of the particle velocity,acceleration factor,and initialization method of the particle swarm on a particle swarm optimization(PSO)algorithm based ona parallel calculation are analyzed to determine the improved direction.With an expanded upper limit of the particle velocityand regenerated particle swarm in the iterative procedure,the results showed that the modified PSO algorithm improved thesimulation times with a fitness value greater than 7500 m(a 95%improvement).The convergence of the algorithm was shownto be faster and the number of calculations remained the same.Thus,the overall efficiency of the solution was improved.

关 键 词：潜艇 自航式声诱饵 鱼雷 粒子群优化算法 

分 类 号：TJ631.5[兵器科学与技术—武器系统与运用工程] U666.75[交通运输工程—船舶及航道工程]