Research On Path Planning Of Industrial Robot Based On Improved RRT Algorithm

With the continuous development of sensing technology and control technology,as well as the implementation of the "Industry 4.0" strategy and the "Made in China 2025" policy,the manufacturing industry has been promoted to transform to the direction of intelligence and automation.Industrial robots are a key technology in the field of intelligent manufacturing,and are now widely used in welding,assembly,palletizing and spraying,to improve the reliability and safety in the industrial production process,and to liberate the labor force.Path planning is a supporting technology in the field of industrial robot research.An appropriate path planning algorithm can ensure the safety of industrial robots and improve the working efficiency of industrial robots.In this paper,the collision detection of industrial robots is carried out to establish the configuration space.On the basis of RRT algorithm,the coverage elimination mechanism is proposed to enhance the search ability of the random tree.The target orientation of the random tree is enhanced by using the target guidance probability.The end strengthening mechanism is added at the end node of the random tree to achieve rapid convergence.The node simplification and path smoothing of the initially generated path are carried out to make the final path tend to be optimal.First,introduce the research background and development status,explain the significance of the research,and analyze the future development trend of industrial robots,the importance of path planning in the application of industrial robots is explained,the research status of path planning algorithms at home and abroad is described,the hot spots and difficulties of path planning research are analyzed,and the research content of this paper is proposed.Then,the kinematics modeling and analysis of the small industrial robot used for loading and handling studied are carried out.On the basis of the parameters of KUKA R540 industrial robot,the improved D-H parameter method is used to establish the mathematical model,and the transformation matrix between the joint connecting rod coordinate systems is constructed to quantify the space relative position relationship.The inverse kinematics equation is solved by analytical method,and the optimal solution is found by the principle of minimum joint change;The correctness of kinematics equations is verified in Matlab,and the workspace is solved and visualized by Monte Carlo method.Secondly,the common collision detection algorithms are compared and analyzed.Combined with the industrial robot model studied in this paper,a hybrid bounding strategy combining sphere bounding box and cylinder bounding box is proposed to simplify the industrial robot model;For the collision detection of spatial cylinders,a progressive spatial position limitation method is proposed,which can realize collision detection for cylinders with different spatial positions and judge the collision type.Thirdly,establish the configuration space,compare the configuration space with the workspace,establish the configuration space through the uniform sampling method,establish the self collision database and obstacle collision database of the industrial robot,record the collision information,map the collision points to the configuration space,establish a complete cloud of collision points and conduct boundary visualization.Finally,the shortcomings of RRT algorithm and its derivative algorithms are improved by adopting a bidirectional search strategy,adding a coverage rejection mechanism to the random tree to enhance the search ability of the random tree,introducing the target guidance probability and guidance method to enhance the target convergence during the random tree search,adding a bidirectional end strengthening mechanism to the two random trees to achieve fast convergence,and performing node simplification and path.The final path is smoothed so that the final path converges to the optimal one.The data are compared and analyzed by twodimensional and three-dimensional simulation experiments to verify the superiority and practicality of the improved algorithm.