基于流形学习的多轴加工—机器人操作协同轨迹规划与奇异点规避算法OA

In multi-axis machining robot machining systems,the coupling between non-orthogonal motion chains and continuous rotation axes resulted in inherent kinematic redundancy and parameterized singu-larity in the configuration space,which cannot meet machining requirements.Therefore,a multi-axis ma-chining robot operation collaborative trajectory planning and singularity avoidance algorithm based on manifold learning was designed.A robot head coordinate system was established,and a multi-axis machi-ning robot operation trajectory kinematic model was constructed through robot rotation transformation and origin translation.Robot motion trajectory data were mapped to manifold space,an equivalent uncon-strained optimization objective function for trajectory planning within the manifold learning framework was constructed,and then the trajectory planning objectives under robot kinematic constraints were estab-lished.By using the matrix of the effect of robot operation speed on the end velocity,non-full-rank singu-larity in the machining robot operation trajectory could be avoided,thereby meeting the consistency re-quirement between the robot operation collaborative trajectory and the trajectory expectation.The final ex-perimental results showed that the linear velocities of the robot operating the X-axis,Y-axis,and Z-axis fluctuated within the ranges of 0～0.10 m/s,-0.15～0 m/s,and 0～0.5 m/s,respectively;the angular velocities of the X-axis,Y-axis,and Z-axis fluctuated within the ranges of 0～0.05 rad/s,-0.05～0 rad/s,and-0.03～0 rad/s.The planned values of linear velocities and angular velocities were close to the expected values,and the planning effect was good,which played an important role in high-quality ma-chining of robots.