physicsman

How do we can measure the linear speed of a vehicle independantly on the speed
measuring of the wheels?
bm1957
excaza
Can you please re-word your question to something understandable? Are you trying to measure it from the car but not the angular velocity of the wheels? You could use an anemometer (wind speed measurement), though that would obviously be influenced by the wind. You could use a GPS system to get a rough estimate of your speed, I know some commercial GPS units include this.
egnorant
Time the vehicle over a measured distance...do math.
Use this..http://www.therangerstation.com/tech_library/Speedogear.html

Bruce
physicsman
QUOTE (excaza+Aug 2 2008, 07:26 PM)
Can you please re-word your question to something understandable? Are you trying to measure it from the car but not the angular velocity of the wheels? You could use an anemometer (wind speed measurement), though that would obviously be influenced by the wind. You could use a GPS system to get a rough estimate of your speed, I know some commercial GPS units include this.

sorry if my question didn't clear enough,

I want actually to measure the linear velocity of the vehicle not the angular speed of the wheels,
sure I thought to measure it by measuring the wind speed, but it's not speed enough to use that way like the plane,
and I thought about using the GPS, but it will be coastly something,

anyway, I want to use it in the ABS application,
here it's required to measure the angular velocity of the wheels and the vehicle speed to achive this formula (1 - Wv/Ww) or to detect the wheel slipping while suddenly braking.....

I know how to measure the wheel angular speed, but I need to know how to measure the linear velocity of the vehicle....

thanks..
bm1957
QUOTE (physicsman+Aug 3 2008, 09:47 PM)
sorry if my question didn't clear enough,

I want actually to measure the linear velocity of the vehicle not the angular speed of the wheels,
sure I thought to measure it by measuring the wind speed, but it's not speed enough to use that way like the plane,
and I thought about using the GPS, but it will be coastly something,

anyway, I want to use it in the ABS application,
here it's required to measure the angular velocity of the wheels and the vehicle speed to achive this formula (1 - Wv/Ww) or to detect the wheel slipping while suddenly braking.....

I know how to measure the wheel angular speed, but I need to know how to measure the linear velocity of the vehicle....

thanks..

Now it sounds like you want to convert the angular velocity of the wheels into linear speed of the vehicle?

That should be simple... v=rw (v is the linear velocity, r is the radius of the wheel, w is the angular velocity)

I think the most useful to use for r will be the measurement from the centre of the axle to the ground when the tyre is loaded.
physicsman

no I don't need to convert the any speeds ...

anyway I got a way to measure that using the accelerometer...

dose anyone know the physics of the accelerometer??

thanks...
excaza
QUOTE (bm1957+Aug 4 2008, 03:37 AM)
Now it sounds like you want to convert the angular velocity of the wheels into linear speed of the vehicle?

No, (I'm assuming) he wants to measure the speed of the vehicle using a different method in order to compare it to the data he's getting from the wheels.

There are plenty of ways to do so, as stated earlier by a few people. Radar, GPS, anemometer (you can use them at low velocities too), you can even use high-speed video if you have something to use for scale.

QUOTE
dose anyone know the physics of the accelerometer??

The most basic accelerometers measure the deflection of a calibrated weight and convert it into an electrical signal.

Here's something I found on google that explains one of the more common types.
bm1957
QUOTE (excaza+Aug 4 2008, 01:19 PM)
No, (I'm assuming) he wants to measure the speed of the vehicle using a different method in order to compare it to the data he's getting from the wheels.

Yep, I get the relevance to wheel slippage now. Converting angular velocity to linear speed wouldn't be much use!!!

QUOTE
The most basic accelerometers measure the deflection of a calibrated weight and convert it into an electrical signal.

Here's something I found on google that explains one of the more common types.

I think the easiest way to convert accelerometer data to linear speed would be to use v=u+at for small intervals of time. Obviously you would need some way of calibrating at u=0 when the car is stationary. Inertial guidance systems might be another place to look...
cjameshuff
You could integrate the output of an accelerometer, calibrating it to zero when the vehicle is at rest. Look at the accelerometer and gyroscope sensors available here:
http://www.sparkfun.com/commerce/categories.php

They've even got some boards with full IMU packages.
Note that gravity is acceleration, however, so you will have to take into account the orientation of the vehicle...if it is on a hill, it will be accelerating forward/backward/sideways even if it is sitting still. Could be used as a faster, short term input augmenting a GPS receiver, though...and GPS modules are also available through SparkFun.
Enthalpy
I confirm that only an accelerometer delivers a measurement fast enough to be used in an ABS as a complement to wheel angular speed measurement.

Though I doubt the first ABS, who had no ESP, used several kinds of measurements. But accelerometers and gyrometers are welcome for the ESP, and this will be the next task after the ABS.
bm1957
QUOTE (cjameshuff+Aug 4 2008, 02:58 PM)
You could integrate the output of an accelerometer

That was my first thought, but wouldn't it require plotting the data from tha accelerometer and fitting a curve so that you had something to integrate?

Sounds like more effort than it's worth to me. You've got discrete data, so why approximate it with a continuous curve, only to re-quantise it (albeit with infinitesimal strips) when you inetgrate?

Or am I missing something?
Enthalpy
An electronic integrator, in analogue technology, with an op amp, a resistor (or two) and a capacitor. Or rather a low-pass used as an integrator at frequencies beyond its cutoff.

The ADC would be after the integrator.
bm1957
QUOTE (Enthalpy+Aug 4 2008, 05:38 PM)
An electronic integrator, in analogue technology, with an op amp, a resistor (or two) and a capacitor. Or rather a low-pass used as an integrator at frequencies beyond its cutoff.

The ADC would be after the integrator.

Cool, that sounds pretty interesting. Just past my level in electronics by the sounds of it
Dabeer
Why not sonar? They're using it for cruise-control proximity warnings, I believe... Fire a pulse of sound at the ground, and detect the phase change to determine your speed.

There may be problems with this that I'm not aware of yet, but this seems simpler than the accelerometer solution...
cjameshuff
QUOTE (bm1957+Aug 4 2008, 10:57 AM)
Sounds like more effort than it's worth to me. You've got discrete data, so why approximate it with a continuous curve, only to re-quantise it (albeit with infinitesimal strips) when you inetgrate?

Or am I missing something?

Integrate the sensor data numerically. No need to fit to a curve you have an integral for, though doing so for a few points of input at a time can give a more accurate approximation...doing so for two points at a time gives you a trapezoidal approximation, for example, and Simpson's method takes it further, approximating intervals with a quadratic equation. And analog integration is likely a waste of time...analog integrators are tricky to design and build, and there's things like component tolerances and the non-ideal behaviors of real-world components to worry about...not to mention things like changes in component values with ambient temperature and humidity.

http://en.wikipedia.org/wiki/Numerical_integration

Note that any acceleration-integrating system would be unstable and inaccurate in the long run, and would require something to compare against to weed out small errors that would otherwise accumulate. Inertial systems and GPS somewhat complement each other in this...the inertial system gives far faster and more accurate short-term results, while GPS gives less accurate short term results but is immune to long-term drift. If all you need is short-term tracking, you might do without the GPS by just re-calibrating the system's orientation every time the car's still or moving at a constant rate on a level road.
bm1957
QUOTE (cjameshuff+Aug 5 2008, 08:31 PM)
doing so for two points at a time gives you a trapezoidal approximation,

Is that not exactly what I suggested?

It wouldn't strictly be integration though. (AFAIK, numerical integration is a method for approximating the area under a curve, This was kinda my point, there is no curve yet. We have a series of trapeziums, so no need to approximate!) Just a number of discrete data points, with a new piece of data for each pair of points. (v=u+at; u and v being the pairs of points.)

Again, I have no experience of analog inegrators so can't really comment on that
Enthalpy
To the electronics engineer, the sonar is clearly more complicated than the accelerometer. And it may depend more on the surroundings : soil texture, noise, dirt on sensors.

Integrating with a lowpass filter is stable. The output only has the useful information, that is, short-term integration. The unstable long-term integration is removed.

As the integrated signal is much more quiet than the unfiltered output of an accelerometer, it is better to integrate before the ADC - that is, integrate in analogue. The only drawback is that analogue filters with low cutoff frequency take a long time to calm down - longer than a FIR.

In terms of schematics, it's a series resistor followed by a parallel capacitor. And maybe an op amp as a buffer if needed.
cjameshuff
QUOTE (bm1957+Aug 6 2008, 04:39 AM)
Is that not exactly what I suggested?

It wouldn't strictly be integration though. (AFAIK, numerical integration is a method for approximating the area under a curve, This was kinda my point, there is no curve yet. We have a series of trapeziums, so no need to approximate!) Just a number of discrete data points, with a new piece of data for each pair of points. (v=u+at; u and v being the pairs of points.)

You have a series of point samples of a continuous signal. You can approximate the original signal using that data as a series of rectangles, more accurately as a series of trapezoids, and usually even more accurately as a series of quadratic functions. There are various other techniques for obtaining accurate numeric approximations of the integral, how well they perform depends on the signal and hardware available.

Enthalpy: the signal should be properly filtered beforehand, yes. There should not be any frequency components higher than half the sampling frequency, or you get aliasing artifacts (though you can actually use this to your advantage, undersampling a signal to demodulate a low frequency signal from a high frequency carrier...but this is another subject entirely). This *is* integration (with a damping factor, so only high frequency signals integrate to near 0 over a full period), but not the application of it that I was talking about. Integrating acceleration to obtain velocity with analog electronics is not going to be doable over a long enough time period to output any kind of accurate velocity reading without some rather tricky design to minimize leakages that would cause the integrator to drift slowly without any acceleration. And forget about taking changes in vehicle orientation into account with gyroscope sensors.

Agreed on the sonar. You could get a not-very-useful airspeed sensor, but results from a sonar ground speed sensor would be terribly dependent on the type of ground, variations in shape, motion of the vehicle on its suspension, and sonar requires signal processing that is significantly more difficult to do than handling the output of an accelerometer, and depending on how far you go with designing your own, it gets into driver and amplifier design.
physicsman
thanks guys,

I found the solution that is already implemented in the real cars,
there are two things I found:

1- to use the accelerometer that based on the piezoelectric effect, to measuer
the linear speeds.

2- to build a look-up table which have two inputs, one the value which
the driver pushing the Brakes, and the other value coming from the wheel speed.
The output is the slipping value...

see this while I made a simulation: