| 英文摘要 | With rapid development of robotic technology, large amounts of welding robots are widely applied to mid-thick plates welding. Now, many manufacturers of robotic integrated system in china mainly adopt foreign manipulator in the filed of mid-thick plates welding. Additionally, trajectory planning of midthick plates welding manipulator is not mature enough. Therefore, our study focus on trajectory planning of Weaving welding, fitting of NURBS curve, and interpolation algorithm of NURBS curve. Firstly, trajectory planning of triangular, sinusoidal, spiral and crescentshaped weaving welding along spatial line and spatial circle is studied . To meet actual needs, feedrate of weaving is planned to take control of weaving motion through arc length of weld line. Then, mathematical formulas of weaving curve are derived. Then trajectory planning of weaving welding is proposed to adjust the weaving amplitude, the weaving period, the period of manipulator effector staying on either end. Secondly, NURBS curve’s fitting of welding line of complex shape is studied. The theory of NURBS curve and effects of its parameters on NURBS curve are introduced. NURBS curve fitting is analyzed, especially closed curve. In the basis of existing algorithms for optimizing fairness of curve, NSGA-II is adopted to optimize weights of NURBS curve combined with fairness and shape preseving of curve. Finally, the restrictions of geometric and kinematics are analyzed. Then restrictions of chord error , feedrate, acceleration, jerk are obtained. According to obtained restrictions, the feasible feedrate can be achieved. A new NURBS interpolation algorithm based on bisection search is proposed. To compare the performance of proposed algorithm with other interpolation algorithms simulation is conducted, and result reveals that proposed algorithm is higher precision than other algorithms. Eventually, experiment on interpolation algorithm of complex welding line is conducted in the platform of UP6, and the result also shows that proposed algorithm is effective. |
修改评论