A Hybrid A* Path Planning Algorithm Based on Multi-objective Constraints
Yu Zhao, Yongqiang Zhu, Pingxia Zhang, Gao Qi, Han Xue
Abstract
To provide a safe, smooth and efficient global planning path for nonholonomic mobile robots. Aiming at the problems of traditional hybrid $A^{*}$ algorithm in path planning, such as approaching obstacles, unnecessary reversing and redundant turning, a multi-objective constraint method based on hybrid $A^{*}$ algorithm is proposed. To speed up the path planning, the heuristic function is dynamically weighted and the overall path cost function is designed. The path planning experiments are carried out in ROS (Robot Operation System) simulation environment and actual environment respectively, and the results show that. The hybrid $A^{*}$ algorithm with multi-objective constraints increases the minimum distance between the robot and obstacles by more than 50%, reduces the unnecessary times of reversing and turning, and reduces the total running time by 14.2% on average, thus improving the navigation efficiency of the mobile robot.