This block calculates the empirical variance of its input signal. It is based on the formula (but implemented in a more reliable numerical way):
y = mean( (u - mean(u))^2 )
The parameter t_eps is used to guard against division by zero (the variance computation starts at <simulation start time> + t_eps and before that time instant y = 0).
The variance of a signal is also equal to its mean power.
This block is demonstrated in the examples UniformNoiseProperties and NormalNoiseProperties.
|June 22, 2015||