PDA

View Full Version : CEP and driven line analysis



harrington
22-05-2008, 09:38 AM
Hi, I'm wondering if someone at Racelogic can help me visualize the 3% CEP a little better because I'm trying to figure out if I should even try to attempt doing line analysis across different drivers and different cars.

I am attaching a little MSpaint drawing I made up and if you someone could help shade in CEP along the entire line for the following scenarios, I'd appreciate it. For simplicity we'll just assume the the track width across the entire 90 degree section is 10 meters and use that as scale. Blue line is driven line.

1) If I were to take the the same car on the same afternoon (4 hour window) with the driftbox not shifted at all, and drive the exact same line at the start and end of the 4 hour session, what would the max possible variance be?

2) If today I ran, then I come back a week later and drive the exact same line, does the CEP get any worse?

I guess what I'm trying to figure out is if 3% is the absolute maximum error regardless of how many days pass between compared runs, time of year, etc.

Is 3% typical or is it the rare maximum.. ie. by any chance (I am hopeful) might CEP mode be closer to 1 - 1.5%?

Assuming 3% CEP how fast does it shift across an entire lap or even just a single corner lasting just 5 seconds?

If you guys could use the same graphic and shade in the possibilities across different time windows along with a little description for each it would be useful. Maybe already some case study exists that you could point me to?

Many thanks.

Mike Broadbent
22-05-2008, 02:55 PM
I think you've got a little confused...

The standard positional accuracy of GPS is 3 METERS, 95% CEP. CEP means Circle Error Probable - so in other words, every sample taken will land in a 3m radius circle, 95% of the time. This is standard accuracy for ANY global positioning system if you don't have the US Military codes - the only way to increase it is by using a base station (http://www.racelogic.co.uk/vbox/index.php?option=com_content&view=article&id=27&Itemid=24&lang=en), or if your GPS receiver is equipped with DifferentialGPS.

However, velocity and distance measurements are very accurate because they are measured using the doppler effect of the satellite signals. Read this (http://www.racelogic.co.uk/vbox/index.php?option=com_content&view=article&id=37:gps-accuracy&catid=5&Itemid=2&lang=en) for more info.

So the worst you can espect from a drive-line analysis point of view is a 3m error and it won't be any worse than that from one day or week to another.

harrington
22-05-2008, 05:23 PM
Yes I used the wrong terms because I had read that some weeks ago and forgot.

3m radius circle.. so when I'm in actual fact at the apex, the driftbox can show anything from me being 3m OFF the apex, to 3m DEEP on the apex (running into curbs and gravel, etc.

If this drift is slow enough, then I can just translate the entire lap plot into the map using one X and Y limit of the track correct? I guess that works.

Thanks I'll try it out at the track soon.