Objectives vs. Constraint Satisfaction in Goal-Driven Optimization

Screening is the only optimization method that does not require an objective for at least one output parameter. To use any other optimization method, you must define an objective for at least one output parameter.


Note:  For NLPQL, MISQP, or Adaptive Single-Objective, an objective can be defined for only one output parameter.


An optimization problem is undefined when none of its parameters have an objective defined. An optimization problem can also be unable to satisfy a constraint.

Undefined Optimization Problem

The following figure shows an optimization where none of the parameters have objectives defined. This optimization cannot be run.

 

In the Project Schematic, both the state and the Quick Help for the Optimization cell would indicate that input is required. Likewise, in the Outline view for the cell, both the state and Quick Help for the root node would indicates that input is required. If you tried to update the optimization with no objective set, an error message wouldwe display in the Message view.

Constraint Satisfaction Problem

If the constraints defined for the optimization cannot be satisfied, the single Pareto fronts displayed from the Tradeoff chart represent the feasible points that meet the constraints. The infeasible points that do not satisfy the constraints can also be displayed on the chart. This is a good way to demonstrate the feasibility boundaries of the problem in the output space, letting you determine the boundaries of the design envelope that reflects your preferences. Infeasible points can also be displayed on any chart where the optimization contains at least one constraint.

As demonstrated by the following figure, some combinations of Constraint Type, Lower Bound, and possibly Upper Bound settings specify constraints that cannot be satisfied. The results obtained from these settings indicate only the boundaries of the feasible region in the design space. Currently, only the Screening method can solve a pure constraint satisfaction problem.

 

The objective settings of the input parameters do not affect the sample generation of MOGA, NLPQL, MISQP, Adaptive Single-Objective, or Adaptive Multiple-Objective. However, they do affect the ranking of the samples that generate the candidate designs if the optimization is updated.

In a Screening optimization, sample generation is driven by the domain definition, which is the lower and upper bounds for the parameter. Sample generation is not affected by parameter objective and constraint settings. However, parameter objective and constraint settings do affect the generation of candidate points.

When a constraint is defined, it is treated as a constraint if Constraint Handling in the Properties view for the Optimization cell is set to Strict. If Constraint Handling is set to Relaxed, those constraints are actually treated as objectives. If only constraints are defined, Screening is your only optimization method. If at least one objective is defined, the other optimization methods are also available.

Typically, the Screening method is best suited for conducting a preliminary design study. It is a low-resolution, fast, and exhaustive study that you can use to quickly locate approximate solutions. Because the Screening method does not depend on any parameter objectives, you can change the objectives after performing the analysis to view the candidates that meet different objective sets, allowing you to quickly perform preliminary design studies. It's easy to keep changing the objectives and constraints to view the different corresponding candidates, which are drawn from the original sample set.

For example, you can run Screening for 3,000 samples and then use Tradeoff or Samples charts to view the Pareto fronts. You can use the solutions slider in the Properties view for the chart to display only the prominent points in which you are interested, which are usually the first few fronts. When you run MOGA or Adaptive Multiple-Objective, you can limit the number of Pareto fronts that are computed in the analysis.

Once all of your objectives are defined, click Update Optimization to generate up to the requested number of candidate points. You can save any of the candidates by right-clicking and selecting Explore Response Surface at Point, Insert as Design Point(s), or Insert as Refinement Point(s).

When working with a Response Surface Optimization system, you should validate the best obtained candidate results by saving the corresponding design points, solving them at the project level, and comparing the results.


Release 18.2 - © ANSYS, Inc. All rights reserved.