hAPF-ACO:广义障碍环境下的移动机器人路径规划算法

A hAPF-ACO algorithm for path planning of mobile robots in generalized obstacle environments

研究了移动机器人的路径规划问题。针对二值栅格建模法不能完整和真实反映复杂环境地理特征的问题,并考虑到移动机器人任务执行过程中存在破障工况,提出了广义障碍的定义,进而对广义障碍进行了分类,并给出了各种广义障碍的模糊隶属度函数,建立了广义障碍环境栅格地图模型。基于人工势场(APF)算法和蚁群优化(ACO)算法各自的特点,提出了一种求解广义障碍环境下移动机器人路径规划的混合APF算法和ACO算法的h APF-ACO算法。通过对案例的路径规划仿真实验,验证了广义障碍环境下路径规划问题的研究意义;与基本蚁群算法求解结果的对比表明,所提出的h APF-ACO算法在收敛性、解质量及鲁棒性等各方面具有明显的优越性。

Path planning of mobile robots is studied. In view of the two value grid modeling method can not completely reflect the geographical features of the real complex environments,and in consideration of the possible obstaclebreaking conditions when mobile robots perform some tasks,a new concept of ‘generalized obstacle＇is defined,further,the generalized obstacles are classified,and the corresponding fuzzy membership functions for all kinds of generalized obstacles are given,then the generalized obstacle environments are modeled by using the modified grid method. Based on the characteristics of artificial potential field（ APF） method and ant colony optimization（ ACO）method,a novel hybrid algorithm combining artificial potential field with ant colony optimization,called the（ h APF-ACO） algorithm,is put forward to solve the mobile robots＇ path planning problem in generalized obstacle environments. The simulation experiment with nine path planning cases is conducted. The experimental results reveal the significance of the presented problem of path planning,and also show that h APF-ACO algorithm outperforms the ACO method in convergence property,solution quality and robustness significantly.