Re: ikine problem

[Erik van Oene]

Hi,

I have a few remarks.

1. You excluded the rotational part, which may lead to an unreachable pose for the robot. (one of the two reasons I see often resulting in this warning)
2. did you mix interchange T2 and T1 on purpose?
3. You do not give the ikine function n initial guess,  (second reasons I see often resulting in this warning)

try to alter these two lines:

T = ctraj(T1, T2, 50);   -->       T = ctraj(Tz1, Tz2, 50);

q = Robot.ikine(T);      -->       q = Robot.ikine(T,[0, 0, 0, 90, 0, 0]);

and the code runs fine here.

Erik

Op dinsdag 24 oktober 2017 20:52:17 UTC+2 schreef ...mail.com:

Hello,

I am trying to solve inverse problem for 6DOF manipulator (I am trying to do a kinematics for human arm).

I want to test the trajectory from one point to another but this problem occurs:

Warning: ikine: rejected-step limit 100 exceeded (pose 1), final err

0.81 

> In SerialLink/ikine (line 244)

  In test_workspace (line 35) 

Warning: failed to converge: try a different initial value of joint

coordinates 

To get the beginning position of end effector and the end position I performed fkine method:

Tz1 = Robot.fkine([0, 0, 0, 90, 0, 0])

Tz2 = Robot.fkine([0, 0, 0, 120, 0, 0])

It returned me this:

Tz1 = 

   -0.4481   -0.8940         0      89.4

         0            0              1         0

   -0.8940    0.4481         0     155.2

         0            0              0         1

 

Tz2 = 

    0.8142   -0.5806         0     58.06

         0            0              1        0

   -0.5806   -0.8142         0     281.4

         0            0              0         1

So I assumed that (x = 89, y = 0, z = 155) could be my start point and the end one could be a (x = 58, y = 0, z = 281)

but it appears that something went wrong... Could someone help me with it? Thanks in advance.

P (1) = Link([0  100  0  pi/2]);

P (2) = Link([0  0    0  -pi/2]);

P (3) = Link([0  100  0  -pi/2]);

P (4) = Link([0  0    0  pi/2]);

P (5) = Link([0  100    0  -pi/2]);

P (6) = Link([0  0    0  0]);

Robot = SerialLink(P);

Tz1 = Robot.fkine([0, 0, 0, 90, 0, 0])

Tz2 = Robot.fkine([0, 0, 0, 120, 0, 0])

T2 = transl(89, 0, 155) % define the start point

T1 = transl(58, 0, 281) % and destination

T = ctraj(T1, T2, 50); % compute a Cartesian path

% now solve the inverse kinematics

q = Robot.ikine(T); 

Robot.plot(q);

[Erik van Oene]

Hi,

This is probably a bug in the SE3 class which I fixes a while ago. Although I was not able to find my fix anymore.

does this line works for you?

T = ctraj(Tz1.double, Tz2.double, 50); % compute a Cartesian path

Erik

Op woensdag 25 oktober 2017 19:23:53 UTC+2 schreef p....ail.com:

Hello Eric!

Thank you very much for your support.

I changed the code according to your remarks:

P (1) = Link([0  100  0  pi/2]);

P (2) = Link([0  0    0  -pi/2]);

P (3) = Link([0  100  0  -pi/2]);

P (4) = Link([0  0    0  pi/2]);

P (5) = Link([0  100    0  -pi/2]);

P (6) = Link([0  0    0  0]);

Robot = SerialLink(P);

Tz1 = Robot.fkine([0, 0, 0, 90, 0, 0]);

Tz2 = Robot.fkine([0, 0, 0, 120, 0, 0]);

T = ctraj(Tz1, Tz2, 50); % compute a Cartesian path

% now solve the inverse kinematics

q = Robot.ikine(T,[0, 0, 0, 90, 0, 0]);

Robot.plot(q);

This is matlab output:

>> test_workspace

Error using SE3/interp (line 437)

error in trinterp

Error in SE3/ctraj (line 474)

                traj(i) = T0.interp(T1, s(i));

Error in test_workspace (line 29)

T = ctraj(Tz1, Tz2, 50); % compute a Cartesian path

did you mix interchange T2 and T1 on purpose?

No, it wasn't done on purpose it was just a copy paste error.

Could you please help with it?

Best regards,

[program...@gmail.com]

Yes, it helped!

Thank you very much for your help Eric.

Best regards

[ask2...@gmail.com]

Hi,

This discussion is also very helpful for me too. I am working on RRRP Robot.
Thank you very much Eric and program...@gmail.com for hlep .

Best Regards
Zain

[Peter Corke]

Using RTB10.2 your example

P (1) = Link([0  100  0  pi/2]);
P (2) = Link([0  0    0  -pi/2]);
P (3) = Link([0  100  0  -pi/2]);
P (4) = Link([0  0    0  pi/2]);
P (5) = Link([0  100    0  -pi/2]);
P (6) = Link([0  0    0  0]);

Robot = SerialLink(P);

Tz1 = Robot.fkine([0, 0, 0, 90, 0, 0]);
Tz2 = Robot.fkine([0, 0, 0, 120, 0, 0]);
T = ctraj(Tz1, Tz2, 50); % compute a Cartesian path

q = Robot.ikine(T,[0, 0, 0, 90, 0, 0]);

works perfectly

[rood...@gmail.com]

please how can in find my Mask value ? what is the technique

[Peter Corke]

You have not provided enough information to answer this.  Have you read the documentation for ikine()?