Jumping motion in point to point movement

Question

I have one question about point to point movement.
Let's say the end-effector is placed at $(0,0,0)$ and the goal position is at $(10,0,10)$. The corresponding joint angle value changes $d\theta$ can be computed using an inverse kinematic solver as follows:

$d\theta=J^{-1}*(10,0,10)$

In this case, the computed joint angle value changes will make the robot jump from $(0,0,0)$ to $(10,0,10)$. I tried to prevent this jumping motion by dividing the goal position into $(1,0,1)$ so that the robot can slowly move towards the goal. Will there be other effective methods?

Answer 1

Do you require the end-effector to hit all the via points $(1,0,1),(2,0,2),...(9,0,9)$? Or do you simply want a lower speed as your end-effector swings towards its destination (with no regard for the path traced out by the end-effector)? If it's the latter, you could set an upper limit to the velocity in your motion planner.

If you want to hit all the via points and have a smooth continuous motion, make sure the velocities obtained from the polynomial trajectories at those points are non-zero.