基于改进粒子群优化算法的巡航导弹航路规划

粒子群优化(PSO, Particle Swarm Optimization)算法是继遗传算法、蚁群算法之后的又一种新的群体智能算法,经常用于复杂问题的求解.由于其迭代公式是面向连续空间的,因此更适合解决非网格拓扑的航路规划问题.标准的粒子群优化算法在寻优的过程中容易出现早熟现象,针对这种现象,提出了一种改进的粒子群优化算法.改进算法根据相应的代价函数选择精英粒子和较差粒子,对较差粒子采用了带有动能补偿的速度更新策略,从而避免了寻优过程中的早熟现象;在单个粒子的运动方面引入了最差粒子的失败经验,让群体中粒子有效避开最差解.仿真表明:改进算法在航路规划的应用中具有更强的搜索能力,获得的航路代价在进化代数相同的前提下更小. 

Route planning of cruise missile based on improved particle swarm algorithm

Particle swarm optimization algorithm(PSO) is new type swarm intelligence algorithm after genetic algorithm and ant colony optimization algorithm,which is usually used in solving complex problems. Because its iterative formula is continuous, PSO is more suitable to solve route planning without grid. To the problem of premature frequently appeared in standard particle swarm optimization, improved particle swarm optimization (IPSO) algorithm was proposed. IPSO firstly selected elite particles and bad particles according to relevant cost function,updated velocity of bad particles according to kinetic energy loss of elite particles to avoid premature in search process. Secondly IPSO proposed velocity update strategy with failure experience of worst particles to let particles avoid bad result. Result which use IPSO in route planning of missile shows that, IPSO has better search capability in route planning application and receives smaller cost if iterations are same.