Design Variables

Design variables are model parameters that typically influence performance requirements, component & physical limitations and production-based limitations amongst other factors.

• All single-valued, numerical parameters from models drawn in MagNet, ElecNet or ThermNet are identified as design variables by OptiNet.
• Design variables are one of four types:
  • Constant (variable will not be optimized)
  • Continuous
  • Discrete Step
  • Discrete List of values
• Sensitivity analysis indicates the influence of each design variable on the optimization objectives.

Constraints

Constraints, which represents any type of limitation or specification that a variable is subjected to, guide the optimization process since the optimal configuration will be searched only amongst those devices which satisfy them.

• Constraints are weighted, indicating their importance and need to be satisfied with respect to acceptable solutions.
• Infinite weights are permitted and indicate an absolute constraint that must be satisfied.
• Model-based constraints only require the parameterized model to be evaluated.
• Solution-based constraints required the field solution to be evaluated.
• Common constraints are included with OptiNet. Users can define their own constraints as well.

Objectives

The objectives are the optimized target, which maximizes or minimizes particular aspects of the model as described by an objective function.

• There can be multiple objectives for a problem, each is assigned a weight to indicate their importance.
• Model-based objectives only require the parameterized model to be evaluated.
• Solution-based objectives require the field solution to be evaluated.
• Common types of objective functions are included with OptiNet. Users can define their own objectives as well.

The Goal is the sum of all weighted/normalized objectives and weighted non-absolute constraints that aren't satisfied.