Research On Path Planning Algorithm For Indoor Robots Based On ROS

Mobile robots are a hot topic in current scientific research and are widely used in industries such as catering,mining,and military.Indoor robots are even more visible in daily life,such as catering robots,JD logistics vehicles,and so on.Path planning is a key technology for mobile robots,and path planning algorithms can directly affect the efficiency and safety of robots.This article focuses on the path planning algorithm for indoor robots.Based on the A * algorithm and dynamic window method,a hybrid path planning algorithm is designed,and its effectiveness and practicality have been demonstrated through experiments.Aiming at the problems of traditional A * algorithm such as low planning efficiency and high number of transitions in complex environments,an improved A * algorithm is designed.Firstly,the bidirectional search strategy is introduced into the traditional A * algorithm,and the advantages of the bidirectional search A * algorithm are analyzed and compared;In order to accelerate the search efficiency,the impact of heuristic functions on the A * algorithm was analyzed,and dynamic weights were ultimately selected to balance the node expansion speed;In order to avoid repeated visits to nodes by two-way search and improve the probability of meeting in the middle,the traditional A * algorithm’s 8-neighborhood search method is extended to 24-neighborhood search,and the selection strategy of nodes that can be passed is redesigned;By comparing the effects of polynomial fitting and Bezier curves,a third order Bezier curve was ultimately selected as the final determined path point for smoothing.In order to improve the adaptability of traditional dynamic window methods to the distribution of obstacles,an improved dynamic window method is designed.Firstly,in order to reduce the steering angle at obstacles and increase the driving speed in open areas,the difference between the heading angles at the front and rear times is adjusted Adding a trajectory evaluation function;Then,by comprehensively analyzing the impact of the weights of each part of the evaluation function on trajectory planning,considering planning efficiency and driving safety,the weights of the evaluation function are finally determined.By analyzing the shortcomings of the improved A * algorithm and the improved dynamic window method,and combining the advantages and disadvantages of the two algorithms,a hybrid path planning algorithm is proposed.The global optimal path key point obtained by the improved A * algorithm is used as the local target point of the improved dynamic window method,and the effectiveness of the hybrid path planning algorithm is verified through experiments.Simulate the real physical environment on ROS and Gazebo simulation platforms,and conduct simulation experiments using the turbobot 3 two-wheel differential robot.The path planning and obstacle avoidance experiments were conducted in the built ordinary indoor environment and indoor warehouse environment,respectively.The experimental results show that the hybrid path planning algorithm can avoid all obstacles in different environments to design a global optimal path,while being able to timely avoid unknown obstacles that suddenly appear in the environment.