Skip to content

A comparison between two methods of the nonlinear optimization: Sequential Quadratic Programming and Semismooth Newton

Notifications You must be signed in to change notification settings

tianmingdu/SQP-vs.-Semismooth-Newton

 
 

Repository files navigation

SQP vs. Semismooth-Newton

This is a project using MATLAB to make a comparison between two methods of the nonlinear optimization: Sequential Quadratic Programming and Semismooth Newton.


Directories tree

|-- SQP-vs.-Semismooth-Newton
    |-- tex
    |-- test
        |-- problem_...
        |-- ...
        |-- problem_...
        |-- extras
        |-- test-populator
            |-- src
                |-- data
                |-- main

The main directory SQP-vs.-Semismooth-Newton contains two subdirectories, three Matlab files, one config file and this README file.

The three Matlab files are:

  1. seq_quad_prog.m (The implementation of SQP method)

  2. semismooth_newton.m (The implementation of Semismooth Newton method)

  3. active_set_strategy.m (The implementation of Active Set Strategy)

The two subdirectories are:

  • Directory tex contains the LaTeX files for the bachelor thesis.

  • Directory test contains MATLAB and Java files for the test.

    • The MATLAB files in this directory with the prefix test_all can be used to run all the test files.

    • Each test problem as MATLAB test file can be found in the different subdirectories. Every test problem get his own directory.

    • The directory test_populator contains source code in Java to create all the MATLAB test files.

    • The directory extras here contains other MATLAB files needed for the test or for the results interpretation.


About

A comparison between two methods of the nonlinear optimization: Sequential Quadratic Programming and Semismooth Newton

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • MATLAB 60.7%
  • TeX 26.5%
  • Java 10.0%
  • Smarty 1.4%
  • M 1.4%