Generate normally distributed random numbers for physical modeling
The Random Number block generates normally (Gaussian) distributed random numbers. To generate uniformly distributed random numbers, use the Uniform Random Number block.
The block behavior is the same as the Simulink® Random Number block (except that it generates a physical signal rather than a unitless Simulink signal) and is based on the polar rejection method [1, 2].
You have an option to specify an initial time offset as part of the Sample time parameter. In this case, the block outputs 0 until the simulation time reaches the offset value, at which point the random sequence starts.
Random Number blocks that use the same seed and parameters generate a repeatable sequence. The seed resets to the specified value each time a simulation starts. By default, the block produces a sequence that has a mean of 0 and a variance of 1.
Specify the mean of the random numbers. The default is 0.
Specify the variance of the random numbers. The default is 1.
Specify the starting seed for the random number generator. Output is repeatable for a given seed. The seed must be an integer in the range of 0 to (2^32 – 1). The default is 0.
The value of the time step interval. The default step value is 1 s. To specify an initial time offset, enter the parameter value as [step, offset], otherwise the offset value is assumed to be 0. The offset must be less than the step size.
 Bell, J. R. "Algorithm 334: Normal random deviates." Communications of the ACM. Vol. 11, Number 7, 1968, p. 498.
 Knop, R. "Remark on Algorithm 334 [G5]: normal random deviates." Communications of the ACM. Vol. 12, Number 5, 1969, p. 281.