Monte Carlo

Figure 88: Monte Carlo definition

The Study settings pane allows to define the following options:

• Sampling method

• Parameters

• OLH Manager

The Study parameter definition pane contains the following two tabs:

• The Inputs/Controls tab contains the list of inputs that can be declared as controls.

• The Outputs/Responses tab contains the list of outputs that can be declared as responses.

For both lists, the first column contains the names you gave to each parameter in the Parameters tab.

Sampling method

Select from:

• Optimized LH

• Latin Hypercube

• Random

See Latin Hypercubes for more details on these methods.

Parameters

• Number of runs Use this field to specify the number of runs you want to perform.

• Seed Simcenter Amesim uses a pseudo-random number generator to choose the input parameter values. In such a generator, the initial seed value determines all the values that will be generated. If you do a Monte Carlo execution twice with the same seed value, you get exactly the same results. In order to perform different Monte Carlo execution with the same parameter settings, change the seed value.

OLH in database

OLH in database

When you select Optimized LH as the sampling method, the application indicates whether an OLH with the selected parameters (Number of factors, Number of runs and Seed value) already exists in the OLH database:

The "LED" is green: an OLH with the selected Number of factors, Number of runs (17) and Seed value (1) already exists in the OLH database.

The "LED" is grey: No OLH with the selected Number of factors, Number of runs (17) and Seed value (2) exists in the database.

If no corresponding OLH is stored in the database you can click Select Nearest to set these parameters automatically, as described in the next section.

Select Nearest

Select Nearest

When you select Optimized LH as the sampling method, the Select Nearest button is enabled. This button automatically sets the Number of factors, Number of runs and Seed value to the closest match from among the OLHs stored in the OLH data file. For example, if you specify Number of runs: 16, and the "LED" is grey, this means there is no matching OLH in the database. If you then click Select Nearest, the application sets the value to that of the closest match:

Number of runs: 16. Seed value: 1. LED: grey. No OLH in the database matches these criteria.

Click the Select Nearest button. The application sets the parameters to the closest match found in the OLH database: Number of runs: 17. Seed value: 1. The LED turns green.

If no such OLH exists and you click Start, the OLH will be generated before performing the runs.

Manage OLH Database

Open the Optimized Latin Hypercube Manager:

See The OLH Manager for more details on this facility.

Inputs/Controls

Monte Carlo Controls tab

Figure 89: Inputs/Controls tab

The Name column contains check boxes. If ticked, Simcenter Amesim will make the parameter value vary respecting the following conditions:

• The distribution type set in the Distribution column: Uniform or Gaussian.

• The mean value set in the Mean value column.

• The value stored in the Std deviation/Amplitude, which is:

the standard deviation for a Gaussian distribution

the amplitude of the total range for a uniform distribution

If an input parameter is not ticked as used, the value specified in the Mean value column will be used for all runs.

Output/Responses

The responses list is shown in the figure below.

Figure 90: Output/Responses list

The Name column contains check boxes. If ticked, the value obtained for each run will be stored in the log file and will be available for post-treatments.