1
$\begingroup$

I'm working on getting a solution to the inverse kinematics of a 6DOF Articulated robot with a wrist joint.

I derived the robot's forward kinematics firstly, using these angles $$θ_1 = 60.102$$ $$θ_2 = 65.220$$ $$θ_3 = 40.390$$ $$θ_4 = 70.215$$ $$θ_5 = 40.002$$ $$θ_6 = 30.211$$

Using these joint parameters, I found a total transformation matrix

$a_1 = 0.18;~~a_2 = 0.6;~~a_3 = 0.12;~~a_4 = 0;~~a_5 = 0;~~a_6 = 0;$

$d_1 = 0.4;~~d_2 = 0;~~d_3 = 0;~~d_4 = 0.62;~~d_5 = 0;~~d_6 = 0.115;$

$α_1 = 90;~~α_2 = 0;~~α_3 = -90;~~α_4 = 90;~~α_5 = -90;~~α_6 = 0;$

This is the resulting total transformation matrix:

$$-0.920890~~~0.342692~~~~0.185808~~~~-0.077297$$ $$-0.010244~~~0.455209~~~~-0.890326~~~-0.273985$$ $$-0.389689~~~-0.821795~~~-0.415687~~~0.845690$$ $$0.000000~~~~0.000000~~~~0.000000~~~~1.000000$$

Now, I'm trying to find an analytical Inverse Kinematic solution that will output the original joint angles I used for the robot's forward kinematics using the total transformation matrix I found.

I used this formula to find the first three angles by decoupling the robot. The output of $\theta_1$ is correct but the output of $\theta_2$ & $\theta_3$ are wrong.

$$θ_1 = atan2(y_c,x_c) + π$$ $$θ_3 = atan2(s3,c3)$$ $$θ_2 = atan2((z_c-d_1),\sqrt(x_c^2+y_c^2)) - atan2(a_3s_3,a_2+a_3c_3)$$

KEY

$$x_c = P_x - (d_6*r_{13})$$ $$y_c = P_y - (d_6*r_{23})$$ $$z_c = P_z - (d_6*r_{33})$$ $$s_3 = sine theta 3$$ $$s_3 = sqrt(1-c_3^2)$$ $$c_3 = cosine theta 3$$ $$c_3 = \frac {(x_c^2 + y_c^2 + (z_c-d_1)^2 - a_2^2 - a_3^2)}{(2*a_2*a_3)}$$ $$a_2,a_3,d_1 == robot link parameters, defined earlier$$

Does someone know of another IK analytical method I can use that will get me my original joint angles?

$\endgroup$
2
  • 1
    $\begingroup$ The problem is that for a 6-DOF arm there is not a unique set of angles that arrive at the solution $\endgroup$
    – guero64
    Mar 19, 2022 at 16:04
  • $\begingroup$ can you post the axis assignment of your robot at the home configuration ? $\endgroup$
    – McLovin
    Oct 19, 2022 at 16:40

1 Answer 1

1
$\begingroup$

You don't say which DH parameter conventions you are using, so it will be hard for users to help you. And there are actually three different DH parameter conventions.

But what might be going on is that they are both valid solutions, but with different configurations. A 6 DoF arm can have many different joint configurations which lead to the same end-effector pose. Here is a simple 2D example:

two different IK solutions

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.