Home
Optimisation in Feko
Feko offers state-of-the-art optimisation engines based on generic algorithm (GA) and other methods, which can be used to automatically optimise the design and determine the optimum solution.
Defining an Optimisation Goal
An optimisation goal defines the request type to be optimised and the goal it will attempt to achieve.
Focus Processing Options
Specify what operation should be performed on the selected goal.

Welcome
Release Notes
See what's new in the latest release.
Get Started
The Feko Getting Started Guide contains step-by-step instructions on how to get started with Feko.
Examples
The Feko Example Guide contains a collection of examples that teaches you Feko concepts and essentials.
Introduction to Feko
Feko is a comprehensive electromagnetic solver with multiple solution methods that is used for electromagnetic field analyses involving 3D objects of arbitrary shapes.
CADFEKO
CADFEKO is used to create and mesh the geometry or model mesh, specify the solution settings and calculation requests in a graphical environment.
POSTFEKO
POSTFEKO, the Feko post processor, is used to display the model (configuration and mesh), results on graphs and 3D views.
EDITFEKO
EDITFEKO is used to construct advanced models (both the geometry and solution requirements) using a high-level scripting language which includes loops and conditional statements.
Feko Solution Methods
One of the key features in Feko is that it includes a broad set of unique and hybridised solution methods. Effective use of Feko features requires an understanding of the available methods.
Optimisation in Feko
Feko offers state-of-the-art optimisation engines based on generic algorithm (GA) and other methods, which can be used to automatically optimise the design and determine the optimum solution.
- Optimisation Workflow in CADFEKO
  The workflow for setting up an optimisation in CADFEKO is explained.
- Launching OPTFEKO (Windows)
  Use the most suitable option for launching OPTFEKO.
- Launching OPTFEKO (Linux)
  Use the most suitable option for launching OPTFEKO in Linux.
- Command Line Arguments for Launching OPTFEKO
  Use the command line arguments to pass additional information to OPTFEKO when it is launched.
- Optimisation Methods and Stopping Criteria
  The duration and accuracy of an optimisation depends on the selected optimisation method and stopping criteria.
- Optimisation Parameters
  Specify the variables to alter during the optimisation run.
- Optimisation Masks
  An optimisation mask is a graphical approach to define an optimisation search. More complex scenarios are handled by a mask, where the goal shape is visually known, for example, the desired bandpass and bandstop regions of a filter.
- Defining an Optimisation Goal
  An optimisation goal defines the request type to be optimised and the goal it will attempt to achieve.
  - Structure of an Optimisation Goal
    Each part of the definition in an optimisation goal serves a specific purpose and should be correctly understood and applied for the desired optimisation outcome.
  - Optimisation Goal Types
    Select an optimisation goal consistent with the requested output.
  - Focus Processing Options
    Specify what operation should be performed on the selected goal.
  - Goal Operator
    Specify how the focus is compared with the goal objective and the goal operator.
  - Goal Objective
    Choose between a single value or range of values defined by a mask to which the required output is compared with.
- Global Goal: Combining and Weighting of Multiple Goals
  Specify the method for combining multiple goals. A weight or importance factor is assigned to the combined goal and optimised according to the combination type, for example, the average of the goals combined for the specified data.
- Optimisation Using .PRE File modifications
  Use optimisation with modifications made in the .pre file in EDITFEKO.
- Optimisation Solver Settings
  Make adjustments to the optimisation settings for a more computationally efficient solution.
Feko Utilities
The Feko utilities consist of PREFEKO, OPTFEKO, the Launcher utility, Updater and the crash reporter.
Description of the Output File of Feko
Feko writes all the results to an ASCII output file .out as well as a binary output file .bof for usage by POSTFEKO. Use the .out file to obtain additional information about the solution.
Feko Application Macros
A large collection of application macros are available for CADFEKO and POSTFEKO.
Scripts and Application Programming Interface (API)
CADFEKO and POSTFEKO have a powerful, fast, lightweight scripting language integrated into the application allowing you to create models, get hold of simulation results and model configuration information as well as manipulation of data and automate repetitive tasks.
Appendix
Reference information is provided in the appendix.

Focus Processing Options

Specify what operation should be performed on the selected goal.

The following processing steps are common to all goals.

No processing: Where the focus is non-complex, no processing steps are required. In order to consider the focus directly, the No processing option is provided.
$x \to x$
Real/Imaginary/Magnitude/Phase: Selects a specific component of a complex focus type. For an array, the complex component of each array element is taken, delivering a non-complex array.
$x \to Re (x); x \to Im (x); x \to Phase (x); x \to Mag (x)$
Unwrap: Unwraps a phase component. For a phase array, the whole array is considered in the unwrap process. This operator is applied directly after selecting Phase.
$x \to unwrap (x)$
Absolute value: Takes the absolute value. For an array, the absolute value of each element is taken.
$x \to | x |$
Average/Minimum/Maximum: Finds the average, minimum or maximum value of an array. This has no effect on a single value.
$x \to ave (x); x \to \min (x); x \to \max (x)$
Normalise: Normalises to the largest value in an array. For a single value, 1 will be returned.
$x \to \frac{x}{\max (x)}$
Log: Takes the base-10 logarithm. For an array, the base-10 logarithm of each element of the array is taken. This operator is only available for non-complex values or arrays.
$x \to \log_{10} (x)$
Offset: Adds a specified non-complex value. For an array, the value is added to each element of the array. This operator is only available for non-complex values or arrays.
$x \to x + n$
Scale: Multiplies by a specified scale factor. For an array, each element of the array is multiplied by the scaling factor.
$x \to n x$
Exponent: Applies an exponent. For an array of values, the exponent of each value in the array is taken.
$x \to x^{n}$
Undefined: When a processing step is modified and the step becomes invalid, the processing step reverts to an Undefined state. Delete or redefine all Undefined steps before applying the changes to the goal.