Solver Settings options

Performance Settings

Presolve: This option reduces the size of the mixed-integer linear programming model before attempting solution generation. Presolve uses various techniques to simplify the problem, such as deleting unused variables and substituting fixed variable values with constants. The resulting model is typically less complex and will most likely solve faster than the original model. When setting this value in a scenario, select “Presolve” as the field from the Network Optimization Options table and select "Do not use any", "Use all" or "Use all except integer reductions".

Cut Strategy: Defines the limits to which potential solutions are excluded from analysis. Cuts are additional constraints that cut off parts of the linear program (LP) solution space, but not solutions of the mixed integer program (MIP). With an aggressive cut strategy, a greater number of cuts are generated than with a conservative cut strategy. This reduces the number of nodes to be explored, but there is a time cost associated with generating the cuts. When setting this value in a scenario, select “Cut Strategy” as the field from the Network Optimization Options table, and select the appropriate strategy.

Heuristic Strategy: Defines the limits to which new solutions are explored in the solution. A heuristic is an algorithm for finding feasible solutions to the problem. When setting this value in a scenario, select “Heuristic Strategy” as the field from the Network Optimization Options table and select the appropriate strategy.

Cutoff Value: Defines a target value at which the optimization will stop and return the solution matching or surpassing the specified value. When you set a cutoff value, Network Optimization only searches for solutions that are better than the cutoff value, effectively reducing the number of nodes to process and improving performance. Use this option if you already know a feasible solution to the problem. Entering an inappropriate cutoff value can result in an infeasibility. To set this value in a scenario, select “Cutoff Value” as the field from the Network Optimization Options table and enter the numeric cutoff value.

Tolerances

Feasibility Tolerance: Defines the boundary at which a solution is considered feasible or infeasible. This tolerance is applied to constraints. For example, you have defined a constraint:

5x + 3y <= 10

and the feasibility tolerance is set to 0.01 (1.0E-2). If 5x + 3y <= 10.01, the constraint is satisfied. However, if the feasibility tolerance was set to 0.001 (1.0E-3), this value, 5x + 3y <= 10.01, would not satisfy the constraint.

In a poorly scaled model (typically has both extremely large and extremely small numeric values), if you encounter infeasibility, you may want to increase the feasibility tolerance by a small amount. To set this value in a scenario, select “Feasibility Tolerance” as the field from the Network Optimization Options table and enter the numeric value representing the tolerance.

MIP Tolerance: Defines the boundary at which binary or integer values can be considered flexible non-integer values. In other words, this is the amount by which an integer variable can differ from an integer and still be considered feasible. For example, if the MIP tolerance is set to 0.01 (1.0E-2), a value of 0.01 is considered to be 0, whereas a value of 0.02 would not be considered 0. To set this value in a scenario, select “MIP Tolerance” as the field from the Network Optimization Options table and enter the numeric value representing the tolerance.