Unicycle: Dancing in the kitchen.

[Motherfuckersito]

kool.........congratulations!!­!! :)

[RolloM41]

This is a cool video. But this is not dancing at all!!!!!!! By definition Dancing involves Music and usually the dancer has some interaction with the song.

Cool Video, but this is not Unicycle Dancing. Try changing the name of the clip cause your outa step.... :)

[reinoks]

Hey !

I like the "one wheel balancing" idea ;D

Did you position optimized sensors?

What do you think of wheel inside a wheel balancing robot idea? By wheel inside a wheel i mean wheel positioning like in ring thought ring magic trick that positions are

[eicosanoid]

Very nice! Do you have a project page for this? What kind of microcontroller are you using, and what is your control algorithm - I notice you have quadrature encoding on the reaction wheel. I have a lot of questions!

[McGuywer]

The gyro doesn't pass to me as very efficient correctional force for this particular case. It's a rotational force (straight forces pointing in all directions), so after you do all the summing of these forces with respect to the contact point with the floor, the resulting useful force will be weak.

You get the max. force out of the gyro if you place your rig in the center of the gyro. But then you'd have to make it a "rope walker", as half of the gyro goes below floor level.

[DANVELVILL]

"Sum of forces with respect to a point" = torque. The torque of the upper wheel depends on its mass, diameter and angular acceleration, not its position.

However, this torque must counteract the falling torque, which basically depends on the mass of the whole robot and the position of its centre of gravity. That's why balancing my robot is not easy.

A "rope walker" is much easier, but no fun at all, you can just put the centre of gravity under the rope and then you need no wheel at all.

[scottgdietz]

Really awesome feat of engineering. Question! What do you call the wheels suspended in the air? Are they gyros? Thank you, S.G. Dietz

[DANVELVILL]

I would just call it "inertia wheel".

A physical gyro also contains an inertia wheel, but it keeps spinning fast in a single direction preventing the movement of the axle by means of coriolis forces.

My inertia wheel works in a different way, accelerating clockwise or counterclockwise to counteract the torque of the gravity, that tries to make the robot fall.

[DANVELVILL]

About disturbances, lateral balance is critical, I haven't measured how many degrees it can recover from but certainly not more than 2-3.

When I balance it on a soft surface, like a bed the ground wheel moves forward and backwards in a straight line because the friction prevents turning around the vertical axe Z. However on a hard surface the friction is low and it eventually turns around Z. If this happens at certain forward speed then the centrifugal force makes it fall (see long balance vid)

[DANVELVILL]

yes, the robot is standalone, but it has a bluetooth serial port for remote debugging and parameter tunning from a laptop.

It has 3 axes acelerometer and 3 axes piezo gyro and I use complementary filtering to calculate the real angles. That way the robot starts controlling when it detects it is properly vertical.

The controller scheme is identical for both axes: inner PI loop controlling torque to the motors and outer (slower) PI loop adjusting the angle references.

[jvpommeren]

Well done! It seems you've embedded the controls onboard of the robot. Sweet!

Can it also withstand disturbances? What kind of controller are you using for the lateral plane?

[tqpolyvac]

When are you going to do the unicicle for humans? ¿Cuándo harás el uniciclo para humanos?LOL ¡Es realmente muy genial!It´s really impressive...

[DANVELVILL]

Search for "eunicycle".

the human controls the lateral balance with his arms and the electronics take care of the forward balance.Isn't it great?