So I made a Chung Method Solver: Triathlon Forum: Slowtwitch Forums

So I made a Chung Method Solver

Jun 16, 12 13:50

Post #1 of 9 (2171 views)

http://www.wandertechnologies.com/parameter-estimation/

Basically it parses Garmin .tcx files and uses an extended version of the Chung method to estimate CdA and Crr. Only inputs you need to add in are total mass of you and the bike, the air density, and what lap to use from the file. From there it will run and spit out the estimates. Full instructions are on the site.

It's still pretty rough, hacked it together over the past couple of days, but try it out and let me know what doesn't work. I've set up a bug reporting form so I can actually fix things as they come up.

___________________________________________________

http://www.wandertechnologies.com

Quote Reply

Re: So I made a Chung Method Solver [willmcg4132] [ In reply to ]

Quote | Reply

ZackCapets

Jun 16, 12 13:59

Post #2 of 9 (2155 views)

I'm assuming it requires power data or will it work without?

__________________________

I tweet!

Quote Reply

Re: So I made a Chung Method Solver [ZackC.] [ In reply to ]

Quote | Reply

willmcg4132

Jun 16, 12 14:09

Post #3 of 9 (2140 views)

No it needs wattage data. The general idea of the method is that you have a known input to the system (wattage from a power meter) and a known output (position from a GPS), so you can use that information to estimate what is in between (the system).

If you use a Garmin as a headunit, you have all of that information in one neat place. Newer Garmins use .fit files, which I have not yet messed with, so safest bet is to upload your file to Garmin Connect, export it as a .tcx (.gpx doesn't include power data), and then use that file.

___________________________________________________

http://www.wandertechnologies.com

Quote Reply

Re: So I made a Chung Method Solver [willmcg4132] [ In reply to ]

Quote | Reply

Tom A.

Jun 16, 12 14:32

Post #4 of 9 (2110 views)

willmcg4132 wrote:

I don't want to rain on your parade or anything (I think it's cool that you did this) but if you're relying on GPS-based speed for this then you might have less than desirable results. The largest source of error in the method is the speed measurement, and IME the GPS-based speed values are just too "noisy".

I'll give it a shot with some of my field test files and see how it matches up :-)

http://bikeblather.blogspot.com/

Quote Reply

Re: So I made a Chung Method Solver [Tom A.] [ In reply to ]

Quote | Reply

willmcg4132

Jun 16, 12 14:38

Post #5 of 9 (2098 views)

Apparently I have to wait a bit before SourceForge will post it, so PM me your email and I'll send you the file. I would be very interested to see how the tests compare.

___________________________________________________

http://www.wandertechnologies.com

Quote Reply

Re: So I made a Chung Method Solver [Tom A.] [ In reply to ]

Quote | Reply

willmcg4132

Jul 4, 12 12:19

Post #6 of 9 (1920 views)

Thanks to some help from the guys over at the wattage group, I've managed to make a few improvements to the tool, should be much less buggy, a little more flexible, and in the next couple of days will support direct import of .FIT files. Download the newest version and try it out.

http://www.wandertechnologies.com/parameter-estimation/

___________________________________________________

http://www.wandertechnologies.com

Quote Reply

Re: So I made a Chung Method Solver [willmcg4132] [ In reply to ]

Quote | Reply

willmcg4132

Jul 8, 12 17:00

Post #7 of 9 (1683 views)

http://www.wandertechnologies.com/parameter-estimation/

Another update:
-Fixed issues people were having when not using the wheel speed sensor. (wheel speed still going to be more accurate)
-Improved data cleaning, dealing with zeros or non-existent data
-Fixed some issues with importing .FIT files
-Fixed an issue with wind-direction math, seems to be changing results a small amount.
-Have for the most part built out new functionality, but is disabled for now, more details on the website.

As always, let me know if you run into any problems and I will try to get them sorted out as quickly as I can.

___________________________________________________

http://www.wandertechnologies.com

Quote Reply

Re: So I made a Chung Method Solver [willmcg4132] [ In reply to ]

Quote | Reply

willmcg4132

Jul 22, 12 8:32

Post #8 of 9 (1505 views)

http://www.wandertechnologies.com/parameter-estimation/

Newer version is up.

I've enabled the power curve and simulation tabs, but they are still pretty rough. Basically you can make a power curve file, from individual files or a folder. Once it is made, you can also add to it as you go. With that and your parameters (mass, rho, cda, ect.) you can use the simulation tab to estimate fastest possible performance over a course. It really works best on climbs, because braking and turning aren't accounted for. It's also pretty rough, and I haven't done anything to speed it up yet. But it's interesting albeit slow.

Also, notably, thanks to some help from users, I have tracked down an error in the GPS heading calculations that was throwing off the way wind was accounted for. Basically the old atan2 vs atan vs arctan debacle. All headings were being reported in the 1st quadrant. Should be dealt with now.

As always, let me know if you have any problems, the first 2 tabs should be pretty stable by now, the 2nd two are questionable.

-Will

___________________________________________________

http://www.wandertechnologies.com

Quote Reply

Re: So I made a Chung Method Solver [willmcg4132] [ In reply to ]

Quote | Reply

willmcg4132

Sep 5, 12 12:04

Post #9 of 9 (1244 views)

Ok, rather large update:

http://www.wandertechnologies.com/parameter-estimation/

I've added in 'day-to-day functionality'. Basic data viewing, metrics, and fitness tracking. There is a diary of daily training which you can sort a number of different ways to quickly see things like your best 20min efforts, longest rides or highest average powers. You can save notes for a ride. Everything, including data, notes, and the data to build plots is saved in .csv and .txt files so that you can use the data for anything else you want easily.

Another little cool thing is altitude correction. A friend had an old paper/formula floating around with a formula to normalize wattage to sea level. We thought that might be interesting after he spent a week in Colorado and was having trouble looking at his training rationally because he had spent most of his life getting used to seeing sea level numbers. Not something that is going to change anyone's life, but it's interesting I think.

Other than that, a few bugs have been fixed in the VE portion, and the simulation tab has had some issues dealt with.

I am still working on more complete documentation, but everything there is at least touched on in the tutorial on my site.

The major change you will notice is that I am now charging $10 for it. I am not trying to get rich or anything, and the purchase will include all updates to the existing modules for life, but at this point, there are alot of hours in this project.

So... let me know what you think.

___________________________________________________

http://www.wandertechnologies.com

Last edited by: willmcg4132: Sep 5, 12 15:10

Quote Reply

Poll

Triathlon Forum

Our Partners

Poll

Triathlon Forum

Our Partners

Newsletter