• aditya mehrotra.

chip updates: standing and sitting without IK [updates]

So this post won't be great because WIX KEEPS DELETING IT BECAUSE THE BLOG FEATURE IS SO ANNOYING SOMETIMES. Sorry, this is the third time I've had to write this.

But basically. What we're trying to do is get the robot to move its legs with JUST the joysticks getting it to stand and sit down in full-user-operated mode.

How we're going to achieve this is by reading the joystick value (a value from -1 to 1) and adding that value to the shoulder and knee joints and keeping the hinge joint exactly the same. I'm going to try KP of 0.5 for now and increase if need be.

At first we added the same value to each of the legs and it worked but the legs move in the wrong directions because we need to reverse some motors.

We just need to fix from this above code, first all the legs were referencing the frontLeft leg instead of themselves so we need to fix that. And then on the left side the knee joints were reversed and on the right the shoulders we moving backwards. This is as simple as changing if we add or subtract the increment.

And VOILA that works. All legs are moving "up" or "down" now. The only thing we want to change is the scaling factor because the legs are moving slowly. Once we scale all the increments by 2 and we scale the knees individual by another 2, we are getting the expanding trajectory we want! Here is the result.

So as you can see the legs are moving in a pretty good way here. But we can't get a robot to stand using this trajectory because we need to move the shoulders before we move the knees control form two joysticks or this happens.

So we're going to finesse the code a bit. And I'll post the final code and settings here. We're also going to increase KP to 0.1! We also added a KI of 0.5 and made the shoulder and legs individual controlled. After some human error and practice, here is what we got.

Sooo I've been playing with this sitting and standing thing for a while and I've reached some conclusions.

(1) KP=1.0 and KD=0.5 seem to be working OK for the control system at the moment.

(2) There is some drift and the slop is NOT great if I'm honest but we can't be sure of anything with that yet.

(3) These feet do NOT have enough grip there's just no grab on the floor which is causing the robot to be very unhappy and slip and etc so we need to fix that.

(4) We need to put thread-locker on the screws securing the lower-legs.

That's what I got for now. Hopefully tomorrow I can get a creative way to find sticky feet and we can re-try this thing.

#robotics #omgrobots #chipupdates #testing_with_humans

3 views0 comments
© copyright 2019 | aditya mehrotra