Singularity Avoidance And Motion Planning Method For H-beam Plate Welding Robot For Construction

H-beams are widely used in industrial construction because of their reasonable strength-to-weight ratio,strong bending resistance and ease of construction.With the rapid construction of cross-sea/valley bridges,highways and industrial plants,the requirements for the stability and load-bearing capacity of large H-beam sections are increasing,and the welding quality of the tendon plate directly determines its mechanical properties.At present,the welding work of H-beam reinforcement plate mainly relies on manual completion,which is inefficient,harsh environment and poor stability of the weld,seriously affecting the quality of products.As the most effective means to solve these issues,welding robots provide stable welding performance while reducing production costs and increasing production efficiency.In view of this,This research is based on the 6-DOF welding robot IRB1410,combines the characteristics of H-beam welding process,and proposes a solution and optimization algorithm that meets the technical requirements for the singularity and motion planning problems that occur during the pick-and-place and welding of the bar plate,and obtains a better H-beam welding path and process after in-depth research.The main research elements of this paper are as follows:(1)A robot locomotion matrix based on the D-H algorithm was formulated:the position and posture satellites of the robot in 3D space were developed,and the chi-square transformation matrix of the posture and position of any arbitrary stiff object in space and its inverse matrix are derived in theoretical manner;The forward and inverse robot model was constructed based on the D-H coordination frame,solved the conversion formula of positive and negative kinematics,and the correctness of the procedure was verified on the IRB1410 robot;the simulation of robot dynamics was accomplished by Robotics Toolbox,and the robot workspace was calculated by the Monte Carlo algorithm.(2)A "singular separation + improved damping inverse + velocity planning" method is proposed: the robot velocity Jacobian matrix model is established,and the singular avoidance problem is transformed into the problem of correcting the singularity factor to zero.An enhanced target-oriented optimization technique is proposed to accomplish the speed planning of the welding robot,the end speed tracking error of the welding robot in the singular domain;Analysis of this method is carried out through MATLAB simulation and experimental comparison study on IRB1410.The findings suggest that the proposed solution has the characteristics of good real-time,simple process and higher efficiency,which can provide the theoretical foundation for the subsequent robot research.(3)Studied the tendon pick-and-place robot obstacle avoidance path planning and its optimization algorithm: analyzed and studied the commonly used obstacle envelope method,and completed the analysis of the establishment of the collision detection model of the welding system;combined with the basic RRT algorithm,proposing variable step coefficients based on target direction to enhance path seeking performance and decrease path size;introduced the obstacle influence range and robot centripetal force to improve the obstacle avoidance algorithms,The modified simulated annealing algorithm is introduced to smooth and improve the generated paths to obtain the optimal obstacle avoidance paths.experiments to contrast SAVRRT with conventional RRT operator on IRB1410,which proves that the improved path planning algorithm runs fast,consumes less resources and has good real-time performance,and achieves simultaneous obstacle avoidance of the robot end and body.(4)A non-dominated ranking genetic algorithm with time optimal and acceleration smoothing as the objective function is proposed: the initial trajectory curve is fitted by five times B-sample interpolation function;to improve the robot’s robustness and efficiency,the traditional NSGA algorithm with "premature" convergence and poor global search capability is proposed.In order to improve the robustness and efficiency of the traditional NSGA algorithm,which has problems such as "premature" convergence and poor global search ability,three improvement operators are proposed for trajectory optimization,namely,ranking group selection based on fitness,crossover based on target gravity and adaptive neighborhood variation,which always maintain the diversity of the population during the iterative process,The inbreeding phenomenon is avoided,the growth ratio of good individuals is enhanced,the reduction of inbreeding is achieved,and further improve the algorithm’s merit-seeking ability and convergence speed.Finally,the correctness of the trajectory curve is demonstrated with the help of the built H-beam plate clamping and placement production environment.