Bounded distribution
You can create bounded distributions to avoid undesired behavior when samples come from the ends of a distribution tail. For example, you might specify demand that has a normal distribution with mean of 5 and standard deviation of 2. In this case, some samples from the distribution will be negative and will result in nonsensical values and potentially instability in the model.
In these situations, you can specify a lower and upper bound as optional parameters to the distribution that will act as bounds. The bounded distributions and parameters are:
Distribution | Parameters |
---|---|
Bounded Normal (BN) | Mean Std Dev Min (default = 0) Max (default = INF) |
Bounded Exponential (BE) | Mean Max (default = INF) |
Bounded Lognormal (BL) | Mean Std Dev Min (default = 0) Max (default = INF) |
When you apply bounds to the distribution, the graph reflects these parameters:
In the event of a sample that is outside the bounds, the min or max of the bounded distribution is used. With many samples, this results in a slight change to the population mean and standard deviation (in case of a normal distribution). From a statistical standpoint, resampling is not recommended. You could get a sample of -0.01, then resample to get a value that was at the other end of the distribution.
The new distributions and parameters for bounded distributions are available on all columns that support distributions:
Table | Column |
---|---|
Production Policies |
Fixed Order Time Unit Production Time Fixed Order Cost Unit Production Cost |
Site Sourcing Policies Customer Sourcing Policies |
Fixed Order Cost Unit Sourcing Cost Fixed Order Time Source Lead Time Unit Sourcing Time |
Transportation Policies |
Shipment Period Variable Transportation Cost Return Trip Time Load Time Unload Time Load Unit Time Unload Unit Time CO2 |
Production |
Quantity Time Between Orders Order Time Due Date Occurrences |
Process Steps |
Unit Process Time Fixed Order Process Time Average Changeover Time |
Customer Demand |
Quantity Due Date Order Time Occurrences Time Between Orders |
Shipments |
Quantity Order Time Due Date Occurrences Time Between Shipments |
Shipment Routes | Time |
Work Center Changeovers | Time |
Last modified: Wednesday May 15, 2024