FB_GaussianNoise (Function Block)

Generates normally distributed (Gaussian) noise using the Box-Muller transform.

The output follows a normal distribution with the configured Mean and StdDev. Values beyond ±3σ occur roughly 0.3% of the time. This models real sensor noise more accurately than uniform white noise.

Note

The block is FINAL and must be instantiated, not extended. Use FB_init to set fMean and fStdDev at declaration time.

Extends: FB_SoComponent

Properties

Mean

Type: LREAL

Gets or sets the mean.

This is the centre of the distribution, the average value the noise oscillates around. For pure noise centred on zero, set this to 0.

StdDev

Type: LREAL

Gets or sets the standard deviation (σ).

Controls the spread of the noise. Roughly 68% of samples fall within ±1σ of the mean, 95% within ±2σ, and 99.7% within ±3σ. A larger value produces more spread-out noise.

Methods

FB_init

Inputs

Name	Type	Description
`bInitRetains`	`BOOL`	if TRUE, the retain variables are initialized (warm start / cold start)
`bInCopyCode`	`BOOL`	if TRUE, the instance afterwards gets moved into the copy code (online change)
`fMean`	`LREAL`	Mean of the distribution.
`fStdDev`	`LREAL`	Standard deviation. A larger value produces more spread-out noise.

Run

Generates the next Gaussian noise sample using the Box-Muller transform.

Two uniform random samples are drawn per cycle to produce one Gaussian sample. If either sample is zero it is replaced with a small epsilon to avoid a logarithm of zero.