Abstract: Aiming at the problem that the traditional path planning algorithm of Automatic Guided Vehicle (AGV) lacks consideration of actual working conditions,an improved A* algorithm based on the combination of reward and punishment mechanism and narrow multi-bend path avoidance strategy was proposed.The path selection of A* algorithm was optimized and smoothed by designing reward and punishment mechanism.The AGV path efficiency was evaluated in real time by combining with the movement of AGV.Combined with the artificial potential field method,an avoidance strategy for narrow multi-bend routes was proposed to reduce the speed loss caused by turning during AGV operation.By comparing the simulation and experimental results in different scales of random environments,it was verified that the improved A* algorithm had a great improvement in reducing the path time and the number of inflection points,and also had a good effect on improving the path smoothness.

Key words: path planning, A* algorithm, reward and punishment mechanism, artificial potential field method, automatic guided vehicle

摘要： 针对自动导引车(AGV)的传统路径规划算法缺少对于实际工况考虑的问题,提出一种基于奖惩机制与狭窄多弯路线避让策略相结合的改进A*算法。通过设计奖惩机制优化A*算法的路径选择并进行平滑化;结合AGV的运动情况,对AGV路径效率进行实时评估;结合人工势场法提出针对狭窄多弯路线的避让策略,减少AGV运行过程中因转弯造成的速度损失。通过在不同规模的随机环境下的仿真和实验结果对比,验证了改进A*算法在减少路径耗时、拐点数量方面有很大的提升,对于提升路径平滑度也有较好的效果。1.浙江理工大学机械工程学院
2.哈尔滨工业大学机电工程学院

关键词: 路径规划, A*算法, 奖惩机制, 人工势场法, 自动导引车