The MathWorks unveils new genetic algorithm, direct search toolbox for MATLAB

The MathWorks today introduced its new Genetic Algorithm and Direct Search Toolbox for use with MATLAB. The Genetic Algorithm and Direct Search Toolbox extends MATLAB by providing new methods to address additional problem types in the area of optimisation.

Engineers and scientists are constantly looking for approaches to find optimal solutions, perform trade-off analysis, balance multiple design alternatives, and quickly incorporate optimisation methods in their algorithms and models. The Genetic Algorithm and Direct Search Toolbox helps them achieve these goals by making optimisation technology more accessible through graphical user interfaces and structured command line tools.

"The easy-to-use graphical interface and comprehensive application programming interface has helped to enhance our implementation process and direct search methods," said Dennis A Dean II, senior programmer, biomathematical modelling unit, division of sleep medicine at Brigham and Women`s Hospital/Harvard Medical School.

"I can now concentrate my efforts on model building, which is the heart of our group`s work, and assessing what the optimised parameters mean biologically to the problem at hand." With the new Genetic Algorithm and Direct Search Toolbox, The MathWorks continues to expand the range of problems MATLAB can address. The toolbox serves as a central access point for the tools required to use genetic and direct search algorithms. Users can learn from and build customized optimisation routines as needed.

Additional key features include:

* Graphical user interfaces and command-line functions for quickly setting up problems, setting algorithm options, monitoring progress and refining optimisation;

* Genetic algorithm tools with numerous options for creation, fitness scaling, selection, crossover and mutation;

* Direct search tools for defining mesh size, polling technique and search method;

* Functions for displaying run-time visualizations and other custom controls; and

* Automatic generation of MATLAB code for reuse or automating routines.

"With MATLAB and its toolboxes, we provide an integrated environment for optimisation, including a programming language and framework that help to quickly define model components, a deep set of optimisation methods and algorithms, and tools for post-processing and analysing results," said Russ Minkwitz, product manager at The MathWorks.

"With the Genetic Algorithm and Direct Search Toolbox, users now have an option to address problems that are difficult to solve with traditional optimisation techniques, including problems that are not well defined, noisy, difficult to model mathematically, or discontinuous."