基于遗传算法的多车体机器人路径规划方法

Mulei-Body Robot Path Planning Method Based on Genetic Algorithm

摘要:

带拖车的移动机器人系统是一种复杂的多车体系统,在不同的转弯角度下所需的路径宽度也不同,其路径规划是个难题；在给出带拖车轮式移动机器人稳态和暂态过程运动学方程的基础上,推导得出了多车体移动机器人在最大转弯情况下所需的最大路径宽度,并将之定义为等效尺寸；应用遗传算法,提出了基于等效尺寸的多车体移动机器人的路径规划方法；首先通过等效尺寸的多步障碍扩张,对障碍物的临近区域和狭窄通道进行可行性约束的分析；然后应用遗传算法进行全局的路径规划,将路径的长度和宽度综合在适应度函数之中；在一定程度上,多车体移动机器人的路径规划问题得到解决；文章通过建立三维网格化地图,对障碍物的临近区域和狭窄的瓶颈通道区域进行了可行性约束的描述,遗传算法在较复杂的环境中实现了良好的避障,并得到了相对较短的有效路径。

Abstract:

Tractor-trailer wheeled mobile robot is a complicated system, which is composed of a tractor and multiple trailers. As the path width changes with the different turning angles, as its path planning is complicated accordingly. On the base of presenting the kinematics equations of the tractor-trailer mobile robot, the maximum path width is deduced when the robot runs at the maximum turning angle. And this path width is named as equivalent size. Then genetic algorithmic is applied for the path planning method of tractor-trailer mobile robot. Firstly, the barriers are enlarged step by step with equivalent size. Then the vicinities around the barriers and the block neck zone are analyzed and labeled. The fitness function is defined according to both the path length and path width. Finally, the genetic algorithm is adopted to solve the whole path planning problem.In this paper, through the establishment of a three-dimensional grid map of the immediate area of obstacles and narrow bottleneck channel region are described feasibility constraints, genetic algorithm in the more complex environment to achieve a good obstacle avoidance, and has been relatively short a valid path.