Rockwell Automation Arena OptQuest Users Guide - Manual (Page 34)

A RENA O PT Q UEST U SER ’ S G UIDE
30
• •
•
•
•
To include a control that is in an array, select the row in the summary grid and click the
Add Control From Array button. A dialog box will be displayed that allows you to
define a specific element of the control array. The array element will appear in a new row
at the bottom of the grid.
Identifying responses to include as expressions
The Responses node of the OptQuest tree displays a categorized view of all candidate
responses available in the Arena model.
When the responses node is selected, a summary grid on the right displays all the
candidate responses in the Arena model. If a particular category sub-node in the response
tree is selected, such as the Response/Resource sub-node, then the grid on the right lists
only the names of candidate responses for that category.
Responses can be used to create constraint and objective expressions. To include a
response in the optimization problem and to make it available for constraint and objective
expressions, check the corresponding box in the Included column in the Response
summary grid.
To include a response that is in an array, select the row in the summary grid and click the
Add Response from Array button. A dialog box will be displayed that allows you to
define a specific element of the response array. The array element will appear in a new
row at the bottom of the grid.
Note: Responses are outputs of the simulation and cannot be modified in OptQuest.
Specifying constraints
Many optimization models can be formulated without constraints. However, including
constraints (if appropriate), which define a relationship among controls and/or responses,
increases the efficiency of the search for optimal solutions. The Constraints Editor allows
you to add linear or non-linear constraints, which are represented in terms of the controls
that have been selected for optimization. The following expression represents an example
of a budget constraint:
25000 * (MachineCount1 + MachineCount2 + MachineCount3) <= 250000
A linear constraint defines a linear relationship among controls. A linear constraint is a
mathematical expression of linear terms (i.e., a coefficient multiplied by a control) that are
added or subtracted. OptQuest can evaluate linear constraints without running a
simulation. Solutions that violate a linear constraint are discarded by OptQuest.