Aerobic Decoupling Calculation Question

Training
1

Hi all,

How is aerobic decoupling calculated for identified intervals and highlighted sections of a ride?

Thanks!

1 Like
2

The power / HR ratio for the first half of the interval is compared to the second. It only really makes sense to look at this for long intervals where HR lag becomes less of a factor.

phr1 = avg power / avg HR for first half
phr2 = avg power / avg HR for 2nd half
decoupling % = (phr2 - phr1) * 100 / phr1
2 Likes
3

Thanks! I use it by highlighting the interval once my heart rate settles into the interval effort. Typically after 60-90 seconds on a 10 minute interval, for example. Am i thinking of this correctly?

4

That should work but I am also not very knowledgable on how to use decoupling on shorter intervals.

5

Hi @david,

Sorry to revive this topic… but I did not find the answer.

It is just a doubt I have:

Why intervals use average power (according with this answer) instead of normalized watts for the calculation of Aerobic Decoupling in fields?

Thanks in advance!

Edit:

Reading more topics, have this question… Does the 60sec rolling window applied to Decoupling chart in activity power is used for intervals fields too? That’s mean the formula above is not used?

6

Should that subtraction be other way round? If efficiency factor has gone down you want the aerobic decoupling to be positive?

7

Aerobic decoupling only makes sense for longer steady efforts. Avg power and Normalized should basically be the same. Decoupling was introduced many years ago by Joe Friel (or that is at least were it saw it first) and the original definition was to use avg.

8

Thank you @MedTechCD.

Yes I know, and for outdoors long steady rides is almost impossible to be exactly the same although its VI should be close to one.

But, because of my calculations for the fields value in activity chart (not decoupling chart), is why I would want to understand how is calculated in those fields and if these are using that formula why is using avg power instead of NP.

I understand your point, but even with rides with similar steady state, values could be different and in long outdoors rides could happen some issues which make you stop for less than one minute or so…

Also I have read articles of Joe Friel using NP… and actually from others in Training Peaks…

9

I have a Garmin IQ field in beta for calculating aerobic decoupling in real time. It calculates normalised power, but needs to use exponential smoothing averages due to memory constraints of data fields on devices.

image
image361Ă—668 43.4 KB

10

Here is the code that calculates decoupling for an interval. It compares average power / average HR for the first half to the second half.

int mid = powerInterpolator.toOriginalIndex((iv.start_index + iv.end_index) / 2);

double avgW = StatUtils.calcStats(fixed_watts, start, mid, true).getAvg();
double avgHR = StatUtils.calcStats(fixed_heartrate, start, mid, true).getAvg();
if (avgHR > 0.0) {
    double ef1 = avgW / avgHR;
    avgW = StatUtils.calcStats(fixed_watts, mid, end, true).getAvg();
    avgHR = StatUtils.calcStats(fixed_heartrate, mid, end, true).getAvg();
    if (avgHR > 0) {
        double ef2 = avgW / avgHR;
        iv.decoupling = (float)((ef1 - ef2) * 100.0 / ef1);
        //log.debug("start " + start + " ef1 " + ef1 + " ef2 " + ef2 + " decoupling " + iv.decoupling);
    }
}
1 Like
11

Thank you David,

I have to figure out how to implement a custom field with NP instead.

Regards!