一种改进的势场法路径规划算法

发布时间：2018-07-29 08:00

【摘要】：在应用传统人工势场法的移动机器人路径规划问题中,机器人对移动障碍物避障效率较低,路径中存在局部极小点。针对这些缺陷,构建了一种基于势流理论的势场模型。在该模型中,势流理论中的概念与路径规划中的概念一一对应,修正函数解决了势流理论中与路径规划问题的矛盾之处,如点汇处速度无穷大等。为了保证对移动障碍物避障的可靠性,应用茹科夫斯基变换对势场分布及函数进行了改进。模型经改进后,为解决局部极小问题,本文进一步使用了点涡的概念,此后又加入虚拟点源以优化轨迹。最后,讨论了多障碍物势场加权叠加方法。仿真实验中,在多种避障情景下对比了改进前后的势场法。仿真结果表明,改进势场能够引导机器人对移动障碍物进行灵活避障,在避免局部极小点时较传统方法更为有效。
[Abstract]:In the path planning problem of mobile robot using the traditional artificial potential field method, the robot has low efficiency in avoiding obstacles, and there are local minimum points in the path. Aiming at these defects, a potential field model based on potential flow theory is constructed. In this model, the concept of potential flow theory corresponds to the concept of path planning one by one, and the correction function solves the contradiction between potential flow theory and path planning problem, such as infinity of velocity at point convergence, etc. In order to ensure the reliability of obstacle avoidance for moving obstacles, the distribution and function of potential field are improved by using Zhukowski transform. In order to solve the problem of local minima, the concept of point vortex is further used in the model, and then virtual point source is added to optimize the trajectory. Finally, the weighted superposition method of multi-obstacle potential field is discussed. In the simulation experiment, the potential field method before and after the improvement is compared in many kinds of obstacle avoidance scenarios. The simulation results show that the improved potential field can guide the robot to avoid obstacles flexibly and is more effective than the traditional method in avoiding local minima.
【作者单位】：湖南科技大学先进矿山装备教育部工程研究中心;
【基金】：国家自然科学基金项目(61572185,60975069) 湖南省科技计划项目(2014GK3160)资助
【分类号】：TP242

【相似文献】