Skip to content

MATLAB Tool for Multi-Objective Optimization

License

Notifications You must be signed in to change notification settings

ethz-pes/multi_objective_optimization_matlab

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

28 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MATLAB Tool for Multi-Objective Optimization

license - BSD language - MATLAB category - science status - maintained

This MATLAB tool offers different functionalities for multi-objective optimization:

  • Offer a common interface for different solvers
    • Brute force grid search (exhaustive search)
    • MATLAB single-objective genetic algorithm ('ga')
    • MATLAB multi-objective genetic algorithm ('gamultiobj')
  • Offer an abstraction layer to the MATLAB solver
    • Scaling the input variables
    • Generating and filtering initial points
    • Transforming high-level data structures ('struct') to low-level ('matrix')
    • Generating the low-level inputs required by the solvers
  • Allow vectorized and parallel evaluation of the functions
    • Divide the number of points to be evaluated into chunks
    • Evaluate the chunks with parallel computing ('parfor')
    • The points inside a chunk are evaluated in a vectorized way

Mathematically, the following optimization problems are solved:

  • Multiple variables
  • Integer variables
  • Upper and lower bounds
  • Inequality constraints
  • Equality constraints
  • Non continuous objective function
  • Single-objective or multi-objective goals

This tool is developed by the Power Electronic Systems Laboratory at ETH Zurich and is available under the BSD License. The code is also available on the ETH Data Archive.

Example

Look at the example run_example.m which generates the following results:

Adding Solvers

The code is made to take advantage of optimization methods using vectorized evaluation of the objective function. Therefore, it would be easy to add support for 'patternsearch', 'particleswarm', or 'paretosearch'. Adding support for non vectorized solvers ('fmincon', 'fminbnd', or 'fminsearch') is possible but less interesting.

Compatibility

  • Tested with MATLAB R2018b.
  • The gads_toolbox is required (for the MATLAB solvers).
  • The optimization_toolbox is required (for the MATLAB solvers).
  • The distrib_computing_toolbox is required (for parfor loops)
  • Compatibility with GNU Octave not tested but probably problematic.

Author

  • Thomas Guillod, ETH Zurich, Power Electronic Systems Laboratory - GitHub Profile

License

  • This project is licensed under the BSD License, see LICENSE.md.
  • This project is copyrighted by: (c) 2019-2020, ETH Zurich, Power Electronic Systems Laboratory, T. Guillod.