| Description: | Generic objective function |
For a list of properties for this class, see the Objective property reference guide.
Objective objects define objective functions for use in multi-objective optimization.
Each objective object defines an equation representing an objective function to be added to the formulation in addition to or as a replacement for the 'default' minimization objective function.
The 'key' property for Objective objects is Constant which defines the constant part of the objective function (which may be zero). Defining or not, the Constant is the way an Objective object should be toggled in/out of the running Model e.g. by tagging the property with a Scenario. In addition, there are flags to toggle the Objective objects in/out of each simulation phase.
Objective objects have a Sense property to specify whether the goal is to minimize or maximize the objective. The objective function is defined using a linear combination of its memberships which, similar to the Constraint object, can span numerous other objects and properties.
Multiple objectives are combined based on the Priority and Weight properties. When multiple objectives have the same Priority, they are solved together with the Weight values multiplied by their objective function to create a blended total objective for the optimization. When multiple objectives have different Priority values, the optimization problems are solved in order, with the highest Priority objectives solved first. Each subsequent, lower Priority solve includes deterioration limit constraints on the previous objectives already solved for. The Relative Tolerance and Absolute Tolerance govern how much those deterioration limit constraints can be violated.
Note that you can change the priority and weight of the default objective via the Model properties Objective Priority and Objective Weight.
Note: Multi-objective optimization is only supported when using the CPLEX or Gurobi solver.