h a l f b a k e r yA hive of inactivity
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
This idea is for a two-wheeled self-balancing cargo carrying "vehicle," either pulled by a bicycle, or maneuvered by hand. It's not meant to be ridden on, or at least, not by the person pulling/guiding it.
Firstly, the two wheels would either be pneumatic, or tweels. Each wheel is driven by an electric
motor. The battery pack for the motors, and the charger for the battery pack, are incorporated into the vehicle chassis.
A tow arm connects the trailer to the pulling bicycle.
Electronic force sensors detect tension or compression of the arm. The trailer automatically accelerates forwards or rearwards to bring tension/compression on the arm to zero.
Additional force sensors detect if the arm is being pulled to the left or right. The trailer automatically accelerates the wheels in opposite directions to one another, to apply a steering torque to itself to minimize that left/right pull.
It's undesirable for the trailer to push down excessively on the rear of the bicycle towing it, or worse, push upwards. Thus, the trailer must self-balance, and thus neither push upwards or downwards on the bicycle.
Since the distance between the trailer and bicycle is fixed by the length of the tow arm, traditional robotic self-balancing (accelerating the wheels forwards or backwards) won't work.
Instead, it will use an alternative means of moving it's center of gravity relative to the pivot point. Specifically, the wheel/motor assembly is attached to a linear bearing, which is attached to the chassis. The wheel/motor assembly is moved forwards or backwards relative to the chassis by an electrically driven rack and pinion, or (maybe) by a linear motor.
The self balancing system will have at least one digital gyroscope, and one 2D accelerometer, and should be able to use the same type of algorithm as other self-balancing systems (like segways).
Ideally, the balancing system, and the regular motion controlling system (acceleration/steering), can be wholly independent, which would allow the use of two separate processors. Since neither processor will need to do multitasking or parallel processing, this could be much cheaper than using one single more powerful processor. It also makes the system easier to program.
Some models of trailer will have accessories to make them even more useful.
For example, if the bicycle has an electric assist, the trailer might have a DC to DC converter, which would allow the trailer's big batteries to keep the bicycle's small ones charged over long distance trips.
A small petrol-powered generator could keep the trailer's batteries charged for long trips, or off-grid use.
An inverter could allow the trailer to act as a power supply for AC powered electrical devices.
A solar panel array might slow down the rate at which the batteries discharge when the trailer is in use, and gradually recharge them when it's not in use.
[link]
|
| |
It sounds like programming the controller would be a fun
project [+], 'though if I'm not mistaken, the forwards-
rearwards
balancing, at least, can also be accomplished by adjusting
the geometry of the trailer, without the need for a control
system. |
|
| |
My dad kept wanting me to build something like this, so I could pull a trailer behind my bicycle without doing the extra work. I never got him to understand that having an engine running along behind me was the moral equivalent of a motorcycle. |
|
| |
Battery power would be good. I don't like the complex balancing arrangement, though. Either load it to balance or adjust something manually. |
|
| |
I could bun this if it had only one wheel |
|
| |
mouseposture, The problem with adjusting the balance *statically* (by manually moving the cargo) is that, if you balance it on level ground, then go up or down a hill, the load will instantly be unbalanced. |
|
| |
When pulled by bicycle, an unbalanced load will either push the bike's rear wheel away from the ground, reducing traction/handling, or push the bike's rear wheel against the ground, increasing rolling resistance. When pulled by hand, an unbalanced load requires muscle strength to keep the cart from tipping over. |
|
| |
Not to mention, if the cargo isn't securely strapped to the trailer, it may fall/slide off when the trailer tilts as it goes up or down a hill. |
|
| |
bigsleep, well, originally I was thinking of an idea for an electric bike range extender, for things like camping trips -- which would just be a trailer for cargo, with built-in batteries and a cable to keep the bike's batteries charged. Then I realized that the extra weight would be more than an electric bike's small motor could easily pull -- the advantage would be negated -- so the trailer would need it's own motors. It then occurred to me that if you were going up or down a hill with a tall trailer, it would be unbalanced, one way or another (see above paragraphs), so it should self-balance. |
|
| |
baconbrain, while bicycling is good exercise, motorcycles generally get far more passenger-miles to the gallon of fuel than cars... so the moral equivalent of a motorcycle isn't really that bad. |
|
| |
Of course, if you choose to add a petrol-powered range extender to this, it becomes the moral equivalent of a series hybrid motorcycle :) Or at least, it does if it's hooked up to an electric bicycle :) |
|
| |
As for the balancing... I'm quite open to alternative suggestions for self-balancing. |
|
| |
pocmloc, The technique of moving the wheel/motor assembly relative to the chassis could easily be generalized from one dimension to two, allowing a single wheel or ball to support the trailer. |
|
| |
The downside to this (besides added complexity) is that there would be no way for the trailer to turn itself. As a consequence, turning would occur purely due to the pull of the towing bicycle or human. Turning while moving wouldn't be hard, but it would be very difficult to turn the vehicle in-place, due to friction between the wheel/ball and the ground. |
|
| |
On the other hand, a trailer balancing on a single wheel/ball could take corners much faster, with less risk of cargo falling off during turns. |
|
| |
[goldbb] Not what I meant. Of course, what I meant may
be wrong, but, anyway, here it is:
1) The line of the towbar, infinitely extended, intersects
the axle.
2) The load pivots freely around the axle
3) The load's center of mass is *below* the axle
|
|
| |
#3 can be accomplised by either of the following:
3a) suspend the load underneath the axle (prolly
want large-diameter wheels, in this case) or
3b) use ballast so that the center of gravity of the
load+ballast is below the axle. |
|
| |
With this arrangement (correct me if I'm wrong) the load's
weight can't create any torque about the axle, hence,
can't lift or depress the rear wheel of the bike. That
should (again, correct me if I'm wrong) still hold going up
or down hills. |
|
| |
you'd be better off having the controller act on horizontal pulling/pushing and ignoring vertical, just let the hitch deal with that like a regular trailer. |
|
| |
Using large wheels has several problems: It reduces the maximum size of the tractor, since they can't be underneath the chassis. The axle connecting the wheels to each other, and the motor driving/slowing the wheels, will be in the way of the cargo. There's a limit to how high the cargo can be stacked before the trailer flips over. And it looks silly. |
|
| |
Using ballast increases the amount of work that the trailer's motors must do to accelerate, and increases the rolling resistance. |
|
| |
If the tow bar pulls on the axle, and the load pivots freely around the axle then it can be said that no torque will be transmitted from the trailer's load to the towing vehicle. |
|
| |
However, that doesn't mean that no torque will be created by the load -- if the center of mass of the loaded trailer *isn't* centered somewhere on the axle, then whenever the vehicle accelerates or brakes, there will be a torque which will swing the load around the axle. |
|
| |
And if the center of mass *is* centered somewhere on the axle, it will be free to spin around the axle, regardless of the trailer's acceleration -- even if the trailer is still. In this case, a slight touch, or a breeze, will cause it to spin. And if the load isn't very securely secured to the trailer, it will dump off. |
|
| |
AutoMcDonough, that is indeed what I intend -- it should detect if the arm is being pulled forward or pushed backward, and (if it's not a unicycle version) detect if the arm is being pulled left/right... no sensor will explicitly detect if the tow arm is being pulled up or pushed downward. |
|
| |