基于改进A^(*)算法的移动机器人全局路径优化研究

Research on global path optimization of mobile robot based on improved A^(*) algorithm

针对传统A^(*)(A-STAR)算法在移动机器人全局路径规划中存在搜索效率低下、搜索路径不够平滑及不安全等问题,提出一种双向搜索A^(*)算法。在传统A^(*)算法上建立起四邻域及八邻域混合搜索路径机制,设定双向搜索中的虚拟目标点,借助人工势场算法思想,建立双向搜索时搜索点到该虚拟目标点的衰减函数,并在保证最优性的基础上改进评估函数模型以及调试适当的权重值,引入三阶贝塞尔曲线规划新的行驶路径。通过Pycharm平台进行仿真,结果表明:改进后的A^(*)算法所规划的路径长度、搜索效率及路径平滑性等特性都优于传统A^(*)算法,规划出的路径长度较传统算法提高19.61%,搜索效率较传统算法提高66.19%,路径平滑度大幅度提高,全局路径优化效果较为明显。

Traditional A^(*)(A-STAR)algorithm has some problems in global path planning of mobile robot,such as low search efficiency,insufficient smooth and safe search path,and so on.A bidirectional search A^(*)algorithm is proposed.Firstly,based on the traditional A^(*)algorithm,the four-neighborhood and eight-neighborhood hybrid search path mechanism is established,and the virtual target point in the two-way search is set.With the help of artificial potential field algorithm,the attenuation function from the end point to the virtual target point is established.On the basis of ensuring the optimality,the evaluation function model is improved and the appropriate weight value is debugged,and the third-order Bezier curve is introduced to plan a new driving path.The simulation results on Pycharm platform show that the path length,search efficiency and path smoothness planned by the improved A^(*)algorithm are superior to those of the traditional A^(*)algorithm,with the planned path length increased by 19.61%and search efficiency increased by 66.19%compared with the traditional algorithm,the path smoothness is greatly improved,and the global path optimization effect is obvious.