Multi-posture Trajectory Smooth Planning Of Robot Based On Unit Quaternion

With the rapid development and application of robot technology,there is a higher demand for robot performance.Robot motion control technology,as one of the core technologies of robot manufacturing,has an important impact on robot performance.The efficiency and accuracy of robot operation is directly affected by the trajectory planning algorithm in motion control.Thus in this paper,we mainly focused on the study of orientation trajectory planning.Considering the Rotation matrix,Euler,RPY and other methods have the problem of difficult to interpolate,we choose the unit quaternion for orientation interpolation,and using the Unit Quaternion Spline Curve and Quaternion Spherical Linear Interpolation to perform orientation interpolation.Finally,a complete expression of the orientation trajectory curve is obtained by using C2-continuous as a constraint.Since the unit quaternion belongs to the three-dimensional swirling group space,the unit quaternion can not be directly used to program the robot trajectory.In this paper,a quaternion visualization method is proposed.The attitude trajectory is projected onto the unit sphere,and the pose trajectory is constructed,planned and analyzed by using the visualization principle.In order to realize the smooth interpolation of orientation on robot end-effector,Firstly,interpolation curve between adjacent teaching orientations is constructed by Spherical Linear Interpolation.Then determine the first and last point of the orientation transition curve according to the position level which is decided by user.Considering the need of velocity planning,we realized the transformation of the derivative of unit quaternion to angular velocity by using Newton-cotes integral formula.Then the inverse fitting of the orientation trajectory curve expression is carried out by using the third Hermite interpolation.Finally the velocity planning of the orientation trajectory curve is completed and realized the smooth orientation interpolation.The algorithm is verified by Robot toolbox in MATLAB.We analyzed the continuity of the orientation trajectory curve and the variation of the angle velocity and acceleration from the geometric and physical level.At last we achieved multiple orientation smooth transition on Roboworks simulation platform based on the algorithm proposed by this paper.And further demonstrate the rationality of the algorithm at the application level.