The Reserch Of Task Planning Of Arc Welding System With Redundancy

Welding is one of the most important sectors of in modern manufacturing industry. With the development of welding technology, the arc welding system with redundancy is applied more and more widely which can easily achieve capabilities such as avoiding the singularity position and orientation, avoiding obstacle and avoiding the motion of joints surpassing the range. Taking an eleven-axis arc welding system as the research object, this subject aims to research and develop the arc welding task planning system, focusing on motion planning, transition path planning, sequence optimization of the arc welding system with redundancy.With the boat-type welding as the best welding position, motion planning of arc welding system with redundancy has been researched. By the comprehensive consideration of multiple criteria, such as the constraint of joint motion position, welding position, welding torch orientation, motion stability, flexibility, etc, in order to ensure the safety distance between the robot and the environment, the nearest distance criterion based on pseudo minimum translational distance is proposed which can calculate the distance between convex polyhedrons by through linear programming.Then the mathematical optimization model for the motion planning problem is established with the method of incorporating multiple criteria and solved by differential evolution.At last, the efficiency of this method is proved by the motion planning simulation in an eleven-axis arc welding system. The motion of the system is stable and smooth and the links of the robot are far away from the other devices with great flexibility. Besides, welding position and torch orientation satisfy the welding technical requirement, thus the welding quality is guaranteed and the welding efficiency is increased.In addition, transition path planning between two weld seams has been researched. First, taking the STL file of the workpiece as the data source, a lot of position nodes above the workpiece are created in Cartesian space and a network is formed through connecting every two position nodes as an edge which does not collide with the workpiece. Then, A-Star algorithm is proposed for finding a shortest path between the starting points of the current seam and the ending points of the next seam. At last, a method presented to optimize the path. After optimization, the path become much shorter and has fewer nodes, simulation results show that this method is effective.In order to minimize the movement of the positioner, weld seams are divided into clusters according to the positioner's pose when weld seams are in the boat-type welding position. Then the arc-welding sequence optimization problem becomes the generalized traveling salesman problem. Considering the kinematic constraint and the welding deformation constraint, the arc-welding sequence optimization model with the objective of minimizing the welding time is established and solved by the discrete particle swarm optimization (DPSO) algorithm. The DPSO algorithm generates the optimal solution through collaboration and information sharing between particles and has the advantages of simple concept, easy implementation, less parameters to be adjusted, fast convergence speed and so on. Experimental results show that the DPSO algorithm is very promising to solve the generalized traveling salesman problem and the convergent speed of the DPSO algorithm is faster than that of the generalized chromosome genetic algorithm.Finally, the software of arc welding task planning system is developed and integrated into the ROBOLP offline programming simulation platform, including motion planning module, transition path planning module and sequence optimization module. The software function has been tested by simulation, which meets the project requirements.