Design And Implementation Of Robot Automatic Obstacle Avoidance Technology And Path Planning Based On Laser Radar

Since the first industrial revolution,the productivity has been greatly improved.As a new subject category,the research on robot has never stopped.As countries around the world continue to increase their investment in robot scientific research,China has also put forward the plan of "Made in China 2025".Robots are ubiquitous in all areas of life,and as a humanless medium,they are shining brightly during the pandemic.This paper mainly focuses on the robot obstacle avoidance and path planning.The main research content of this paper includes the following aspects:(1)Study the path planning algorithm.Firstly,the framework of the path planning algorithm is explained,and the path planning is divided into global path planning and local path planning.Among them,the global path planning selects Dijkstra algorithm and A-star algorithm,mainly from their implementation principle,simulation process to discuss and compare.DWA algorithm is selected for local path planning,which is explained from sampling,motion model,simulation prediction trajectory and evaluation function.Then,the A-star algorithm is selected for further improvement,mainly from the heuric function,weight coefficient,search node,curve optimization and other aspects.Finally,experiments show that the optimized algorithm is efficient and feasible.(2)Mapping research based on Li DAR.The Lidar M10 is selected as the radar model in this paper,and the relevant parameters are explained.The radar uses the principle of TOF ranging.Aiming at the shortcomings of RBPF principle,a new Gmapping algorithm is derived by improving the proposed distribution function and resampling strategy.Then,the principles of Hector algorithm and Cartographer algorithm are introduced.Finally,by comparing the results of the three algorithms,Gmapping algorithm is chosen as the graph building algorithm in this paper.(3)Experimental simulation of Ackermann car.Firstly,hardware selection was carried out,Jetson Nano was selected as the main control board of the car,and a complete set of hardware and software platform was built.The constitution of ROS system of the car and the function modules needed for drawing and navigation are explained.Finally,2D and 3D mapping and simulation of single point and multi-point navigation are realized by using the trolley platform.The research in this paper shows that the selected global and local path planning algorithms are representative and feasible,and are effective and feasible for the improvement of A-star algorithm.The Ackermann vehicle experiment platform can realize automatic obstacle avoidance and path planning by using the correlation algorithm,and the research on the theory has been effectively verified.