RCnormalise

There's a few people using my script to normalise the data in their RaceChrono CSV files, so I thought I'd start a topic specifically for it. I've released v1.5G today.[P]

Download it here[P]

RCnormalise

Normalises OBD throttle data from 0-100%

Calculates braking effort based on longitudinal acceleration

Estimates current gear based on RPM and speed data[P]

Comments

  • edited October 2009
    DEL
  • The files output for v1.3 and v1.4 should be identical, aside from the different values for braking (same scale, though). Have you tried interpolating with Pinto, then running the output through RCnormalise? Are you sure that file (2.csv) was created with v1.4? It doesn't have the RaceChrono header. Has it been modified since?
  • I've re-run RCnormalise v1.4 on 1.csv to create the normalised CSV and got the same results with Pinto. However, I ran v1.3 on 1.csv too and got the same missing data, so it's nothing to do with the upversion of RCnormalise. I then tried to run Pinto on 1.csv, which failed with a critical error. I'd say there's an incompatibility with Pinto and that original CSV. RCnormalise is quite a dumb script, so it's quite happy to take any data and "fix" it. Pinto may not be so forgiving.
  • edited October 2009
    DEL
  • RCnormalise looks for the column headers to determine which column it needs to adjust. If the column headers are not as the export from RaceChrono is, then it won't be able to make any amendments.
  • edited October 2009
    DEL
  • any chance of a filter for lateral Gs in the next version of normalize or RC itself? I get quit a few very quick erroneous readings of 3-4 Gs as cars go through corners???

    I have been editing them out but that's a pain.
  • Interesting. Do you mean a hard limit for G? So, you could set 1.2G as the maximum your car can cope with and anything above that gets reset to 1.2G? That's dead easy to do, so I can incorporate it. I don't tend to use the G for anything as it seems to be pretty unreliable anyway - it's estimated from the GPS data.
  • exactly. A hard limit field so I can enter a value like 1.5 for tin tops and maybe 2.5 for formula cars.
  • I have had issues with occasional erroneous G force reading like over -4g. There is usually 5-10 of these in the 2 g columns in a 20 minute run.

    If rcn is calculating the braking from the g column and there were erroneous samples wouldn't that affect the entire range and bring braking below what it should be?

    So I decided to pre-edit them before your process. After doing this, the file would not run through rcn.

    Here is a link to the original that works fine and the edited that does not http://drop.io/n1papvs

    Any idea how I can edit these without this happening?

    thanks,

    mr
  • Hi tentenths, I've got your email and seen this. Finding the time to investigate is a bit tricky at the moment. I'll look at it when I can find the time.
  • after the RCNormalise processed the export csv, how to make the braking signal in to Racechrono2avi???
  • edited March 2010
    tentenths:

    Open those two files in Notepad and you'll see immediately that they're very different. Have you changed them in Excel? There's lots of extra quotes in there, which break the way RCnormalise works. It's just a simple script, so it gets confused by non-standard data. Could you do your editing after RCnormalise has been used?

    If you want 1g to be the maximum braking effort, then set that in the app. It will normalise based on that and anything over 1g will be set as maximum braking. There's no need to pre-filter the braking field to remove the erroneous samples, just set your value appropriately.
  • rx8dorifto: I don't think RaceChrono2AVI supports the brake field. RCnormalise supports it for users of other applications that do support that field.
  • Hey Andy,

    I edited them in OpenOffice/win7. In that and Excel/Mac they appear exactly the same but in notepad I see what you mean. I will have to find out why. Thanks for looking in to this.
  • Hi Andy,

    Right now I believe rcn takes "longitudinal acceleration (g)" and converts the negative values in to % and inserts them in to a new columns "brake flag" This works well for me.

    Is there a way to do the same for throttle…like taking "longitudinal acceleration (g)" and converting the positive values in to % and inserting them in to the "throttle position (%)" column? The purpose for this is to get a better simulated bar graph of throttle when you have a file without obd2 data.

    Could this be an easily added option on the current build 1.4 or is there a trick that could use the current build and get this by changing the positive values in the column to negative…then running the script then cut and pasting???

    and trick or workaround appreciated…
  • Hi tentenths,
    [P]
    Yes, it's possible to add this functionality. It's not going to be truly accurate though for several reasons. The problem revolves around the fact that you are measuring acceleration, which is roughly proportional to the torque at the wheels. The torque varies throughout the rev range and decreases when you change up a gear. Not to mention the increasing aero load at speed. It's possible to add an indicative figure, but its accuracy will be poor at best.
    [P]
    What I'm more interested in at the moment is removing some of the noise. There is noise in both the bearing and speed data, which is then used to derive the G data. This results in a large amount of noise in the lateral G data, hence the hard limit bodges I implemented in RCnormalise. I've used the smoothing function in Excel graphs to remove some noise, which works well. I'm also writing a few features in Driving Mirror, the magazine for Bristol Motor Club, which will describe some of the things you can do with simple GPS-based data, RaceChrono and Excel and show some of my workflows. You can subscribe until the end of 2011 for less than 30USD.
  • Andy...I'm sending you an email off forum.
  • Andy, can you email me scans of your articles, would be very useful reading for me :)
  • Driving Mirror is now available online, either via RSS or in iTunes. The first bit of data debugging is a bit basic and doesn't include any of the workflows.
  • Hello everyone!
    Is there a download link for RCNormalise?
  • Looks like the link expired. I've now uploaded the latest version (see download link in first post).
  • dear andylaurence,
    i've tried to download last RCnormalise, but link does not work. have you a new one, or may you send directly to my email?
    ivo.andreini@gmail.com

    thanks a lot for support
Sign In or Register to comment.