I have some questions regarding the power curves smoothing.
I have found on the git hub a file containing all the power curves with different smoothing parameters and I saw in the publication on wind power that a Gaussian filter is used to represent the windspeed distribution within any given hour and the geographical dispersion. The width of this filter is, as written in the article : “s = 0:6 + 0:2 w m=s”.
This formula seems weird to me with this “:”, does it mean : s = 0.6 + 0.2 * w , where w is the windspeed expressed in m/s ?
And in this case, is 0.2 the parameter that is changed between 0 and 0.4 in the 5 differents set of power curve files ?
Which set of power curve is used for the RenewableNinja on the web plateform ?
And how can you efficiently implement such a filter in Python or R ? I saw in the code on github the following code in the convoluteFarmCurve function :
convolver.x = seq(-w, w, 0.01)
convolver.y = dnorm(convolver.x, myMean, mySD)
convolver.y = convolver.y / sum(convolver.y)
smooth_y = filter(y, convolver.y, sides=2)
Does this code smooth the power curve with a Gaussian filter of width s = 0.6 + 0.2 *w ? If so how do we have to chose myMean and MySD ?
Thanks for your answers,