Solver Settings

Tools -> Optimize Model -> Preferences -> Solver Settings
Defines settings for the optimization solver. The values in this dialog box vary based on the Optimization Solver setting in the Optimization Preferences dialog box.
 
For the option:
Do the following:
Settings for all optimization solvers
Name
Displays the name of the optimization solver.
Description
Displays a brief description of the optimization solver.
Verbose
Select to turn on additional diagnostic output.
Convergence Tolerance
Specify the relative tolerance for convergence. The solver will stop when the overall cost has converged to this tolerance.
Maximum Iterations
Specify the maximum number of steps (iterations) to perform. The solver will stop after this number of steps, even if the overall cost has not converged.
Settings for SDI only
Step Ratio
Specify the amount to vary the factor values for each step (iteration). During each step SDI will generate values around the current factor value within this amount of the factor value range. Be sure that the maximum iterations times the step ratio is about one, or more; otherwise, SDI may not be able to use the full range of factor values.
Number of Runs
Specify the number of runs in each step. SDI will generate this number of overall cost evaluations for each step.
Settings for MSCADS options only
Differencing Perturbation
Specify the relative amount to perturb variables during differencing. MSCADS uses finite differencing to compute partial derivatives of the overall cost with respect to the factor values. MSCADS uses this value to compute the perturbed factor values.
Differencing Method
Select the method for computing derivatives using finite differences:
Central - Perturb above and below the nominal value and use the average slope as the derivative.
Forward - Perturb above the nominal value only, and use the slope as the derivative.
Note: Central is slower than Forward, since it requires two additional evaluations of the overall cost instead of one, but it may be more accurate.