.Credibility.Types.TruncatedNormal

Information

A truncated normally distributed uncertainty limited by the minimum and the maximum possible value of the uncertain scalar. See also BaseUncertainty uncertainty for general information.

The additional parameters introduced for the truncated normal distribution are:

• mean: nominal value of the scalar is the mean value of non-truncated normal distribution.
• stdDev: standard deviation of non-truncated normal distribution.

For example, the parameter R of a resistor has the nominal value of 200 Ω, the minimum possible value 190 Ω, the maximum possible value 210 Ω and the standard deviation 3 Ω. This yields the following parameters of truncated normal distribution:

• 200 Ω ± 5 %, σ = 3 Ω ⇒ parameters unitValue ="Ohm", nominal = 200, relTol = 0.05 (and absTol = 0), stdDev = 3 or
• (200 ± 10) Ω, σ = 3 Ω ⇒ parameters unitValue ="Ohm", nominal = 200, absTol = 10 (and relTol = 0), stdDev = 3.

The Modelica Standard Library provides Modelica.Math.Distributions.TruncatedNormal. For more details of the normal distribution or of truncated distributions, see Wikipedia – normal distribution or truncated distribution, respectively.