Precision and drift of GPS position

It's too bad that locomotives don't have a CAN port! The more expensive race car loggers can plug into a car's CAN port and record throttle, brake and steering inputs. Those input can be compared to tachometer output, wheel ticks, and accelerometers to get a pretty complete picture of vehicle response. IIRC, being able to interpret the CAN data stream can require signing an NDA with the auto manufacturer. That was a bridge too far for our project which might be used on a lot of different cars of different ages.
 
I just checked typical braking distances for trains. This about 300-600 meters distance going from 100 km/h to zero.
I think the accurate GNSS receivers (less than 1 meter stand alone) will do just fine. The start of the measurement is more critical at 27 m/s.

Another aspect is certification. The nr's are used to set a maximum allowed speed and I suspect that the railroad companies will ask for "certification".
 
As my collegues just asked me for a GPS solution to avoid cabling, I made further investigation about what this system is suposed to do.
I discovered that this system is not directly involved for certification. It is more an internal tool for the first tests when checking the braking system. But even if it is an internal tool, they still want it to be as precise as possible, and provide repetible measurements.
The real homologation/certification tests is registered with the TELOC3000 (the black box) system, which is connected to various sensors.

Here is the kind of machine I work on:

b66u_01.png


r21_03.png
 
Angelo said:
As my collegues just asked me for a GPS solution to avoid cabling, I made further investigation about what this system is suposed to do.
I discovered that this system is not directly involved for certification. It is more an internal tool for the first tests when checking the braking system. But even if it is an internal tool, they still want it to be as precise as possible, and provide repetible measurements.
The real homologation/certification tests is registered with the TELOC3000 (the black box) system, which is connected to various sensors.

Here is the kind of machine I work on:

View attachment 36515

View attachment 36517
Click to expand...
Nice equipment ;-)

I suspect it to be accurate and repeatable enough. The start of the braking will be the most inaccurate step, but that's more a "group decision" of which moment to take. Teensy wise that seems to be a easy sensor.
 
You must log in or register to reply here.
Back
Top