If one had two or three objective functions, are they executed sequentially one increment at a time, or is one optimization completed and then the next objective function executed?

Objective functions are evaluated at each optimization step. All objective functions are also saved into the MIRaGE project file.  Note that you can also export or import an Objective function (in a directory of your choosing).  The names of the Objective Functions are also reported in the DAKOTA input file (you can view this input file and test it via the Optimization UI).  So if you add multiple functions, all will be used in a single simulation.  In the outputs from a given optimization, there will be a separate column for each objective function as calculated along with the parameters that were varied and exported to DAKOTA.

Thank you for your interest in the MIRaGE software tool suite!

MIRaGE is a new and ongoing R&D area, and as such, our governing entities require some additional information to be gathered and reviewed before distribution.

Acquiring a copy of MIRaGE is a two-step process. The first step will be on the following page when you click the button below. This step is an in-depth questionnaire called the Participant Data Sheet (PDS) and is used to review eligibility.

The second step involves the approval and signature of a Government Use Notice (GUN), which allows entities that either 1) hold an existing government contract with DARPA or another Federal Agency and have a need for the software or 2) have negotiated access to the software through DARPA.

Once the GUN is signed you will receive credentials via an email that allow you to download and install MIRaGE.

If you have issues with the process or questions, please do not hesitate to contact us at help@mirage-software.com.

Click here to start your request

Intelligent Symmetry Engine:

  • Mapping of symmetry to constitutive tensor and optical properties
  • Expert , Guided and Templated build modules
  • Library of meta-atoms and unit cells
  • State of the art CAD with all geometrical primitives and Boolean operations to achieve arbitrary meta-atoms designs
  • User based material assignment to segments of atoms
  • Automatic physical meshing based on material refractive index
  • Pick-and-place arbitrary 3D heterogeneous-unit cell tiling & stitching
  • Arbitrary layered super- and substrates
  • Periodic table of meta-atoms, unit cells, and molecules
  • Supports all space filling tiles in 2D, 2.5D and 3D systems
  • Algebraic expressions for dimensional dependencies

Smart Efficient Multiscale Electromagnetic Simulation Engine:

  • V&V on literature from Microwave-to-optical domains
  • Ability to extract and track diffraction orders
  • Automatic phase unfolding
  • Data processing and extraction of physical properties including: S-parameters, phase, volumetric fields distribution power, pointing vector, field intensity, surface charge density, quiver plots.
  • Exportation of tabulated CVS data and gif temporal time propagation
  • Handling of perturbative nonlinearity
  • User friendly scripting interface for exotic anisotropic material properties including: anisotropic and, dispersive media, gyrotropic media, ferrites with bias fields, graphine, and nonlinear susceptibility
  • Analytic determination of SALC resonant modes pre-simulation
  • Symmetry-based meta-atom periodic table for the design of meta-surfaces and lenses
  • Adjoint based source inversion and masking
  • 1mm2 metalens simulation

Guided flexible Optimization Engine:

  • Incorporation of DAKOTA geometric optimization package
  • Engagement of optimization method via simple mouse clicks in an optimization GUI
  • Canned objective functions and exemplar scripts for user guidance
  • Exposing of all compute quantities including spatial fields vectors and s-parameters to user for defining custom objective functions
  • Automatic propagation of assigned dynamic variables to optimization module
  • Run-time history showing geometric evolution of structure at various stages
  • User ability to upload measured data and use optimization module forensic tool to deduce structural dimensions
  • Incorporation of sensitivity analysis and uncertainty quantification for yielding robust solutions
  • Single and multi-objective optimization with user defined weights
  • Generation and exportation of optimization Pareto-Fronts
  • Incorporation of adjoint formulation and the enabling of exotic inversions including:
  • Source inversion
  • Source masking via active control
  • Target inversion
  • Tensor inversion (numerical homogenization)

Access Request

Access Request

Access Request

Access Request