基于适应度函数的无人船遗传算法航径规划

The Navigation Path Planning of Unmanned Surface Vehicles Genetic Algorithm Based on Fitness Function

摘要:

遗传算法通过适应度函数选取最优的路径,采用了无人船转弯半径来改进适应度函数,实现无人船遗传算法航径规划。考虑到无人船机动性能对航迹平滑性的要求,在初始种群中利用贝塞尔曲线优化方法,将原有的折线路径优化成光滑的曲线路径；在适应度函数中添加曲率判断,以无人船最小转弯半径为约束条件,设置曲线路径的最大曲率,最后通过适应度函数筛选出符合约束条件的光滑路径。仿真结果表明,所提出的方法能获得符合无人船最小转弯半径约束的光滑路径,相比于平滑算法,该方法的曲率更小,收敛速度更快。

Abstract:

Genetic algorithm selects the optimal path through fitness function, adoptsSthe turning radius of unmanned surface vehiclesSto improve the fitness function, so as to realize the navigation path planning that the genetic calculation of unmanned surface vehicles.SConsidering the requirement of trajectory smoothness for the maneuverability of unmanned surface vehicles, the original broken line path was optimized into a smooth curve path by using Bessel curve optimization method in the initialization of population. The curvature judgment was added into the fitness function, and the maximum curvature of the curve path was set with the minimum turning radius of the unmanned surface vehicles as the constraint condition. Finally, the smooth path that met the constraint condition was selected through the fitness function.Simulation results show that the proposed method can obtain a smooth path conforming to the constraint of the minimum turning radius of unmanned surface vehicles. Compared with the smooth algorithm, the curvature of the proposed method is smaller and the convergence speed is faster.