Abstract: In recent years, the optimization algorithm proposed for the traditional artificial potential field (APF) method that was easy to fall into the local minimum problem still had the problems such as low applicability and low computational efficiency. Based on the shortcomings of some improved algorithms, a sampling-based rapidly exploring random tree (RRT) algorithm was innovatively introduced to pre-select a number of temporary target points on a static known map, which avoided the mobile robot from falling into the local minimum value area when the artificial potential field method was used. Meanwhile, the mobile robot carried out the obstacle avoidance strategy of real-time path planning in dynamic obstacle environment. Simulation test results show that the proposed method is simple and easy to implement; meanwhile, it combines the advantages of complete probability and good convergence of RRT algorithm with small calculation and high real-time performance of APF algorithm, which can adapt to the changes of dynamic environment and meet the requirements of dynamic obstacle avoidance of mobile robot.

Key words: vehicle engineering; mobile robot; path planning; artificial potential field algorithm; rapidly exploring random tree

摘要： 近年来针对传统人工势场法（artificial potential field，APF）易陷入局部最小值问题所提出的优化算法依然存在适用性不高、计算效率低等问题，基于部分优化算法的不足，笔者创新性地引入了基于采样的快速扩展随机树算法（rapidly exploring random tree，RRT）在静态已知地图上预先选取数个临时目标点，避免移动机器人在使用人工势场法时陷入局部最小值区域并在动态障碍物环境中进行实时路径规划的避障策略。结果表明：该方法简单易实现，同时结合了RRT算法概率性完备、收敛性良好与APF算法计算小、实时性高等优点，能够适应动态环境变化，满足移动机器人动态避障的要求。

关键词: 车辆工程；移动机器人；路径规划；人工势场法；快速扩展随机树