改进粒子群算法在UUV航迹规划中的应用

Application of improved particle swarm algorithm to UUV 3-D path planning

提出一种改进的粒子群算法,在粒子进化过程中将种群中粒子两个为一对,分成若干对.每次进化之后对每对粒子中的两个粒子进行比较,代价函数小的粒子作为较优粒子正常进化,代价函数大的粒子作为次优粒子,进化时速度变量按一定概率进行变异.通过测试函数对改进粒子群算法性能的验证表明改进粒子群算法具有较好的搜索精度、稳定性及搜索速度.将改进粒子群算法用于无人水下航行器(UUV)三维航迹规划中,仿真结果表明:利用粒子群算法寻找航迹规划代价函数最小点,得到下一规划点,从而实现航迹规划,取得了较好的效果.

A novel particle swarm optimization （NPSO） was proposed. The two for a pair of particles in the population were divided into a number of pairs and compared with two particles for each pair. The particle which shows better cost fitness normal evolution, while the other particle speed variable in a certain probability of mutation. The performance of the particle swarm optimization was tested by several test functions. It is turned out that the NPSO is better than other algorithms in search accura- cy, stability and search speed. Finally, NPSO was used to solve UUV （unmanned under water vehicle） three-dimensional path planning problem, using particle swarm algorithm for path planning cost function minimum points, which gets the next planning points, so as to realize path planning and ob- tain the satisfactory performance.