反舰导弹航路规划的OACRR-PSO算法  

OACRR-PSO Algorithm for Anti-ship Missile Path Planning

作　　者：刘钢[1] 老松杨[1] 袁灿[2] 侯绿林[1] 谭东风[1] 

机构地区：[1]国防科学技术大学信息系统工程重点实验室,长沙410073 [2]海军工程大学核能科学与工程系,武汉430033

出　　处：《自动化学报》2012年第9期1528-1537,共10页Acta Automatica Sinica

基　　金：国家自然科学基金(70571084;61074121);国家部委'十一五'科研汁划(513040404-1)资助~~

摘　　要：为了提高反舰导弹航路规划算法的搜素效率,从几何学角度对航路规划空间进行了研究,在将功能区域概念融入逆向航路规划的过程中发现了功能区域的几何学渐变规律,据此提小功能区域簇作为其物理载体.将功能区域簇引入粒子群优化(Particle swarm optimization,PSO)算法,提出了功能区域簇实吋约束(Operational area cluster reM-time restriction,OACRR)的PSO算法(OACRR-PSO).为了便于表示功能区域簇,采用航路极坐标编码方式.与传统的PSO算法不同的足,考虑到粒子中分量之间的关联性,该算法在优化过程中并不足对粒子的整个速度分量同时进行更新,而是引入一种分步递归进化策略对粒子的分量逐步进行更新.在粒子的更新过程中,使用功能区域簇米实吋限定罻游恢梅至康淖既犯路段В?得算法搜索空间逐步减小,从而加速算法收敛.仿真实验结果表U/J,分步递归进化策略能够非常显著地提高算法的今局搜索.队能,并且算法收敛速度快、稳定性好.In order to improve the search efficiency of path planning algorithm for anti-ship missile, the planning space is researched based on geometric principle. The geometric gradual transformation rule of operational area is revealed when fusing the concept of operational area into the process of converse path planning, hereby, the operational area cluster is proposed to be its physical carrier. By introducing the operational area cluster into particle swarm optimization （PSO） algorithm, a PSO algorithm real-time restricted by operational area cluster （OACRR-PSO） is proposed. To express the operational area cluster expediently, the polar coordinates code mode is adopted in path coding. Considering the relationship between the adjoining vectors of particle, OACRR-PSO does not update all the velocity vectors of particle simultaneously in the course of optimization, which is different from conventional PSO, but updates sequentially by adopting the strategy of sequential recursion evolution. In the course of updating particle, the operational area cluster is used to restrict the position vectors of particle in exact updating area in real-time, which reduces the search space step by step to increase the convergence velocity. Simulation results indicate that the strategy of sequential recursion evolution could improve the algorithmls global search capabilities and the algorithm possesses a better convergence rate and robustness.

关 键 词：反舰导弹 航路规划 功能区域簇 粒子群优化 分步递归进化 黄金分割位置 

分 类 号：TJ761.14[兵器科学与技术—武器系统与运用工程]