h a l f b a k e r y
My hatstand runneth over
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
It's well known that the problem of self balancing can be
solved by making use of powered wheels -- and the
most well known example is the Segway PT.
However, there is one major drawback of this
technique -- since these systems depend on electric
motors, the available torque decreases as speed
increases, which means that above a certain speed, the
vehicle cannot prevent a forward fall. Also, as vehicle
speed increases, the vehicle must tilt further and
further forward -- bringing the front of the vehicle
dangerously close to the ground. In addition, in order to
balance at low speeds, low backlash between motor and
wheels is required, which could result in the gears
binding at high speed due to thermal expansion.
This is an idea for an alternative means of balancing,
one which will work at any speed, without tilting the
vehicle relative to the ground, without /requiring/
powered wheels. And of course, if the wheels are
powered, the system will still keep the vehicle
balanced. Even if the wheels are locked (due to brakes
or transmission), the system should still be able to keep
the vehicle balanced.
Basically, the wheels' axle is attached to an assembly
that is below the rest of the vehicle; this assembly is
attached to the vehicle using linear bearings which
allow it to slide forwards and backwards relative
to the rest of the vehicle.
In addition, there is a linear actuator of some sort
(hydraulic pistons, or a linear motor, or a rotary motor
driving a ball screw, etc.) which moves this assembly
frontward or rearward in a controlled manner.
When the vehicle undesirably tilts forward, the
assembly is moved toward the front of the vehicle to
compensate; when the vehicle undesirably tilts
backward, the assembly is moved rearward to
Determining the tilt of the vehicle should be done by
measuring the distance between the vehicle and the
ground at the front and rear (probably using a pair of
infrared range finders) instead of an accelerometer.
This allows the system to keep the vehicle parallel to
the ground just as easily on sloped ground as it does on
Since the ability to move the assembly frontward or
rearward is /independent/ of the vehicle's speed,
balancing can be done even when the vehicle is moving
In common with a conventional self balancing vehicle, is
a PID controller, and perhaps a gyroscope.
||I like it. Would be helped if people wore helium filled 10 gallon hats too. They would tend to keep people up the right way:)
||One question (not that I know a lot about Segways). In a
Segway, if you tip forward, the wheels do their stuff in
such a way as to (a) provide torque to right you and (b)
bring the wheels back underneath you. Is this about right?
||In your system, the restoring force is provided by a "see-
saw" mechanism (you shift the see-saw's pivot point to
correct the tipping). Will this provide enough righting
force to prevent tipping?
||Why would this system require any less effort to achieve stability than a wheel speed control?
||The system needs an additional motor to drive the stabiliser mechanism. If that motor's power were added to the drive motor power, you'd achive the same effect.
||Using some of the power of the main motor for stabilising makes sense. The motor is already there and it doesn't need the weight and bulk of another mechanism.
||It's also worth mentioning that moving the axle would still require more torque from the main motor - every action has an equal and opposite reaction - so stability would still be speed dependant.
||This is very interesting, but you'd have to change the
driver support mechanism as this would lead standing
people to fly over the handlebars. I do agree with
[Twizz] in that this requires an additional motor
system that would require a lot of power. So it
doesn't seem like a net gain, but interesting enough
to earn a bun. (+)
||//Also, as vehicle speed increases, the vehicle must tilt further and further forward//
||Mmmm, not sure about that...
||The drawback is weight in that an additional motor
of some type is needed.
||MB, yes, two effects; however, since the wheels
of a Segway aren't any more massive than normal
vehicle wheels, it's mostly (a), getting the wheels
back under you.
||If the wheel carrying assembly can travel the full
length of the vehicle, then any amount of
acceleration that this system cannot compensate
for is an amount which would similarly cause a
regular car to flip over.
||Twizz, let's imaging that you've got a self balancing
remote controlled robot, which uses only it's
wheels to balance. Furthermore, you've got it
speeding down a road at it's maximum speed... the
force applied to the motors is exactly matched by
wind resistance. Now, suppose that a sudden
gust of wind gives it a push from behind, and it
starts falling forward. What can the robot do to
avoid smashing it's face into the road? Nothing!
||Now, let's imaging a two wheeled robot which
balances using the principle described in my idea.
It's also moving at it's maximum speed down a
road. It will be level to the ground, and have it's
axle assembly somewhere behind it's center to
balance out wind resistance. A sudden gust of
wind gives it a push forward. What will it do?
Move it's axle frontwards, and /not/ smash it's
front bumper into the ground.
||And, I don't see what you mean when you say that
moving the axle frontward/rearward would need
power from the main motor. Explain?
||MisterQED, this is not for a stand-on vehicle like a
Segway... it's move for something car shaped. So
if there were a driver/passenger, he/she would be
sitting, not standing.
||Remember, unlike a Segway, this kind of balancing
vehicle will remain parallel to the roadway, which
means that one can look straight forwards and see
where one is going -- there's no need to crane
one's neck upwards (one reason why one must
stand to use a Segway).
||And there's no reason why power consumption
would be any greater than in a vehicle that
balances by means of powered wheels.
||Ling, don't forget air resistance! :)
||RS Yes, but how heavy does that additional motor
need to be? Electric motors often have a high
power to weight ratio.
||bigsleep: I'm more concerned with getting a
vehicle up to highway speeds, than putting a
vehicle on a ball.
||The trick is never to use the full power of the motor at full speed. Whether you do this by running one big motor at 90% power, or by having a second motor doing nothing until required is up to you. Either way, you need more power than is required to maintain full speed. A single motor is a lighter, more effecient way to do it.
||Your robot will have a lower maximum speed, as it will be carrying an extra motor and mechanical system to balance.
||Here's another thing: Torque from the drive motor can be altered almost instantaneously, while the moving axle will take time to travel from mid position to either end of it's travel. This introduces a lag into the system. Lag is a disaster for the control.
||The additional mechanical system means that power consumption will certainly be greater than a wheel torque stabilised device.
||I can't see how you expect the thing to remain parallel to the road at speed. Even at 20mph, a runner has to lean forward significantly to remain balanced. What happens if the road is not perfectly flat? This thing will try to keep itself parallel to every pothole and other surface imperfection.
||If the wheel motors aren't necessary for balancing,
then there's no reason why they need to be
electric motors -- a gas engine could power them
just fine. And, if a gas engine is used, then
torque can't be changed instantaneously.
||Lag is definitely a potential problem, but how fast
will the axle need to be moved frontwards or
||The additional mechanical system will of course
use power, but no more power that would
otherwise be consumed by the wheel motors the
purpose of balancing.
||A runner at a constant 20mph needs to apply
rearward force to the ground to balance wind
resistance. To keep his center of gravity "above"
his feet (when considering the combined upwards
and forwards forces the ground is applying to his
feet) he must have his feet behind his center of
gravity. The only way for a human to accomplish
this is to lean forwards.
||A vehicle that balances as described in this idea,
will, when moving at a constant 20mph, also need
to keep it's center of gravity in front of the
location it touches the ground. This can be
accomplished without tilting the whole vehicle,
but instead moving the wheels slightly rearward.
||And if the road isn't perfectly flat... we can use a
moving average of the distance sensors, instead of
the raw data, to determine how tilted the vehicle
is... using the same type of filtering function
that's done with data from the accelerometer in a
more conventional balancing vehicle.
||Thus, the balancing system will ignore potholes
while it's moving, but will adapt to the (much less
transient) changes in road angle as it ascends or
descends a hill.
||Before I forget, there's one huge benefit I forgot to
mention, in case this is being considered for use on
an autonomous robot: Since the vehicle remains
parallel to the ground, we don't need to design the
cameras to look "upwards" in order to see forwards
when moving at speed.