微小型无人水面航行器的全局避障航迹规划

Global Obstacle-avoiding Path Planning of Micro-miniature Unmanned Surface Vehicle

在常见改进遗传算法的基础上,结合微小型无人水面航行器(MUSV)的航行特点,进行变长度实数编码;根据航行的边界约束、避障约束、机动约束、总航路点个数约束以及进行目标点可航性判断来生成初始群体;为了使交叉、变异后的航迹能够避开障碍且能满足航行机动约束,采用相似航路点交叉和优先小范围变异。仿真结果表明,结合MUSV航行特点的改进方法,能够产生适应度较高的初始群体,能够在遗传操作中舍弃不可行个体,从而达到加快收敛速度的效果。

On the basis of usual improved GA algorithm, according to navigating characters of micro-miniature unmanned surface vehicle, variable length floating coding is presented. The first population is built according to the boundary restriction, obstacle-avoiding restriction, maneuverability restriction, number of total swerving spots and judging whether the goal is navigable. To avoid obstacles and satisfy maneuverability restriction, crossing individuals at similar swerving spots and mutating individuals at a small range are adopted. The simulation shows that the improving methods can generate good population, delete bad individuals, and quicken the searching speed.