Segway Physics

[Mansoor]

I was wondering why the segway didn't fall over. at first i thought it was because of the inertia of the rider. but that would mean that the segway would be continuously accelerating. then i thought it was newtons second law. the torque applied by the motor applies an opposite torque to the vehicle keeping it from falling over. but then again that would imply that the motors have to keep applying a force and keep accelerating the thing faster. Then i found a website that modeled the whole thing in matlab, but i think they got the forces wrong. cause they use drag as the force that keeps the segway from falling over, and for all practical purposes there is no drag at the speed the segway goes at.

does anyone know what keeps the segway from falling over.

[little-c]

complicated.... I have no idea how the guys at segway did it, but I would use the change in speed of a motor to shift the angle of the segway. this can be slowing or increasing speed or stopping it altogether.

it takes a lot of thinking. and Im not intrested enough to actuly work it out properly.

[billhowl]

Inverted Pendulum Balancing
http://en.wikipedia.org/wiki/Inverted_pendulum

[Mansoor]

Thanks.

the wikpedia article does explain how to keep a pendulum from falling over. when it is stationary. however it does not say how to maintain an angle when the segway is moving. as with all things that stay in a constant state, the forces and moments on the segway must net to zero for it to maintain a constant condition.

what is the missing force? gravity and normal forces create a moment in one direction. the motors create a moment in the opposite direction when accelerating. drag also creates a moment opposite to gravity.

how do the forces  balance?

[dunk]

if you ignore drag for a moment,
no force is required to keep an object moving at a constant speed so the forces on a moving Segway are exactly the same as a stationary one.

when riding a Segway you actually feel this. you lean forward to accelerate but once moving you are close to straight upright.
when you lean forward the wheels kick forward to maintain the balance, putting you vertical again but now you are moving with constant velocity.

Newtonian physics right?
"""First Law: An object at rest tends to stay at rest, or if it is in motion tends to stay in motion with the same speed and in the same direction unless acted upon by a sum of physical forces."""

obviously drag does provide a decelerating force proportional to the speed.
as a person on a Segway it is quite intuitive to regulate the speed when drag is acting on the Segway. you just lean forward a small bit as you accelerate.
i would imagine such a closed loop would be simple enough on an autonomous platform. to increase speed, increase angle. etc.

is there anyone with relevant fluid dynamics knowledge reading this to tell us if drag is directly proportional to speed?


the more interesting part of the problem are the forces at work while a Segway is accelerating/decelerating.


dunk.

[waltr]

Check this Microchip App Note:http:
//search.microchip.com/searchapp/searchhome.aspx?id=2&q=AN964%20-%20Software%20PID%20Control%20of%20an%20Inverted%20Pendulum%20Using%20the%20PIC16F684&ac=1

This is a simple inverted pendulum.

[Mansoor]

if you ignore drag for a moment,
no force is required to keep an object moving at a constant speed so the forces on a moving Segway are exactly the same as a stationary one.

thanks so much that helped understanding it alot. I haven't had the experience of riding a segway so I thought that to maintain a certain speed the rider must maintain a certain angle. and in that case i could not see how the moment created by gravity is balanced out.

one more question. When i read about the segway it said that there different speed settings. i.e. a certain key is used when the rider wants he maximum speed to be for example 10 miles/hr. in that case how does the segway deal with a passenger who leans forward a little when it already maxed the speed. cause if it does not accelerate a little the segway will not be able to balance the moment created by gravity and the passnger will head face first into the ground

[dunk]

one more question. When i read about the segway it said that there different speed settings. i.e. a certain key is used when the rider wants he maximum speed to be for example 10 miles/hr. in that case how does the segway deal with a passenger who leans forward a little when it already maxed the speed. cause if it does not accelerate a little the segway will not be able to balance the moment created by gravity and the passnger will head face first into the ground

i was over simplifying a little.
if you lean way forward while going full speed on a Segway it readjusts to make sure the centre of mass is directly above the wheels (or a little in front to counteract drag).
so the "handle bars" in this situation would tilt back but the center of mass is still in the same place.

once you have a bit of practice though you stop fighting it and let the Segway point straight up.
this also means your face is not sticking way out in front and the first thing to crash into a fellow Segway rider coming round the corner at you...

i have tried to make a Segway fall over forwards and backwards and as long as it has good traction it really can't be done.
the sensors and drive motors can react more quickly than you can move your weight.

on slippery surfaces all bets are off though.
the worst crash i witnessed was one wheel on a piece of plastic packaging during rapid acceleration. one wheel accelerated quickly while the other just shot out the packaging an stayed put. the rider left the Segway at a suitably comical angle.

if both wheels skid together which is more normal it just dumps you on your arse.


dunk.

[Webbot]

If you are interested then Elektor magazine had a project called the Elektor Wheelie - a seqway by another name.

I can't reproduce the article as it is copyright but you should be able to download the article from their website (it ran over two months ie two editions - plus a later edition for a battery charger) - but you will have to pay a nominal charge (as they need to re-coup the cost of the article!).

Its ATmel based and all schematics, code etc are downloadable. If you don't want to bother with the mechanics you can also purchase the whole thing as a kit (but its not cheap).

Downloading their schematic, article and code may give you some ideas and could be worth the outlay.

[Mansoor]

if you lean way forward while going full speed on a Segway it readjusts to make sure the centre of mass is directly above the wheels (or a little in front to counteract drag).

amassing i gotta try one these things when i get a chance. What does it move to get the center of mass back on top of the wheels? is there some sort of weight inside it that is mounted on a track?

[dunk]

if you lean way forward while going full speed on a Segway it readjusts to make sure the centre of mass is directly above the wheels (or a little in front to counteract drag).

amassing i gotta try one these things when i get a chance. What does it move to get the center of mass back on top of the wheels? is there some sort of weight inside it that is mounted on a track?

nope. it just moves the wheels further forward at the same speed you move your weight forward.

dunk.