With the continuous development of industrial technology,industrial automation has been paid more and more attention to,and robots have become its basic equipment,playing an more and more important role in industrial production.Especially in the field of welding,robots play an irreplaceable role.The robot control system is the core component of the robot,whose performance has an important influence on the overall performance of the robot.Since trajectory planning is an important part of the control system,this paper takes the industrial welding robot as the application background,carries on the research on the industrial robot trajectory planning part.The main contents are as follows:(1)In this paper,cubic NURBS strips is studied,and the interpolation algorithm of Moss differential equation of cubic NURBS is put forward,and the calculation amount is reduced by the method of replacing differential by pre-differential and post-differential combination,and the parameters are obtained quickly by improving the initial iterative formula.By using the parameter fast recursive combined with the prediction-correction method,and through reasonable interpolation preprocessing and reasonable approximate calculation,contour error and feed acceleration control by comprehensive consideration,the dynamic step interpolation algorithm of cubic NURBS curves is finally realized.And the coordinate points under the planned cartesian space are inversely solved to the joint space,and the displacement curves,velocity curves and acceleration curves of each joint are obtained.Finally,the IRB2600-12/1.65 model robot is used as the simulation ontology,and the algorithm is simulated by simulation software.The simulation results show that this algorithm can realize the robot trajectory curve planning as far as possible within the allowable range of error norm.(2)The interpolation function of quintic NURBS for trajectory planning under joint space is studied in detail,which has the properties of ensuring the continuous displacement,velocity,acceleration and jerk curve of each joint.Because the common genetic algorithm can only deal with inequality constraints,which is not ideal for the processing effect of equation constraint problem,and the penalty function method can deal with this kind of problem effectively,an improved genetic algorithm with better performance is proposed by combining the two optimization algorithms.Therefore,this paper solves the problem of robot trajectory time optimization by improving genetic algorithm.Finally,the algorithm is simulated by simulation software,and the simulation results show that the first time-based trajectory is 180.79,the second time is optimized on the basis of time optimization,the impact optimization is 146.51,and the impact is reduced by 18.89% compared with the optimal time trajectory planning.It is proved that this method can get the robot running track which is ideal in both impact and time. |