4 legged robot collapsing

Typically: "How do I... ", "How can I... " questions
Post Reply
patchaaquatica

4 legged robot collapsing

Post by patchaaquatica »

Hi,

I'm trying to build a 4 legged walking robot, and I cant seem to get it to stand by itself, regardless of the joint configuration (IK - both psuedo inverse and DLS, and force/torque). I've fiddled around with joint torque limits and target velocities/positions, as well as varying the mass of links and the body but had no luck.

Here is the scene: https://drive.google.com/file/d/1fxnr9y ... sp=sharing.

I attached child scripts from the hexapod on a similar robot (tweaking them for 4 legs) and managed to imitate the body movement, but on this new one without any child scripts attached it keeps collapsing. Unfortunately I don't have that previous model as reference.


I am also hoping to use distributed control for the legs, in a similar way to the roller-walker, but have the next foot/tip position defined by various dummies. I was thinking initially to link the dummies via a parabolic path, and have the tip follow another dummy along that path. Aside from using the OMPL library, is there another method you know of to generate movement between dummies?

Thanks very much for your help.

coppelia
Site Admin
Posts: 10339
Joined: 14 Dec 2012, 00:25

Re: 4 legged robot collapsing

Post by coppelia »

Hello,

there are several things that are not right or that don't seem right:
  • The joints should be in hybrid IK mode. Otherwise IK calculation won't take place or won't affect anything
  • The masses and inertias seem too small and are not balanced-out. Make sure to carefully read this page.
  • Adjacent dynamic and respondable shapes might be colliding with each other
  • The feet of the robot are not pure shapes. Stability and calculation speed might suffer.
Cheers

patchaaquatica

Re: 4 legged robot collapsing

Post by patchaaquatica »

Thanks for the pointers, I've managed to get a stable configuration now.

Post Reply