If we want to estimate the overall uncertainty in gain measurement of horn antenna, we need to account for all the possible errors such as random effect, proximity corrections, mismatch and multipath effects. And its associated uncertainties as mentioned in the attached files.
As I personally think you have to deal with probability function and it's Gaussian Distribution, when it's math expectation is calculated E(u_1*u_2*...*u_n) = 1/n*[(E(u1)+E(u_2)+...+E(u_n)]
and it's dispersion is calculated D(u_1*u_2*...*u_n) = root_n[D(u_1)*D(u_2)*...*D(u_n)]. Where root_n(x) = √^n(x)=x^(1/n). Other things are not appliable to engineer, but for mathmaticians. It's centered and clear enough. Good Luck Lingasamy!
You need a meaningful samples of specific and independent measurements where you isolated the single error components. Then, you can estimate the single uncertainties and then you can combine them according to the norm. Finally, you must consider if some components of error cannot be considered Gaussian by applying the correction for uniform distribution.
In a simplified way you can make a single and meaningful sample of measurements (independent measurements) which includes all error components. Then, you can make an overall estimate of the uncertainty by estimating the standard deviation of the measurements.
Dear Gifuni,
Thanks for your reply. I have got 10 set of measured values, from which I need to estimate the uncertainty due to proximity effect and multipath interference. Please help me to understand these two terms clearly.
- Badges
- Science topic