Skip to content

Refactored AMOSA, Archived Multi Objective Simulated Annealing, implementation in C based on the code written by Sriparna Saha.

Notifications You must be signed in to change notification settings

amirmasoudabdol/amosa

Repository files navigation

About this version

This version is a Refactored version of the original AMOSA code in order to make the code structure more portable. Several changes have been made:

  1. The AMOSAType type has been defined in order to encapsulate all the necessary parameters of the AMOSA. This will reduce the number of external variables, in addition, it will reduce the possible conflict when it uses within another routine.
  2. The amosaParams type which defines in the amosa_real.c will pass to all the functions during the process.
  3. Functions name are changed for clarification
    • ReadParameters reads the commands lines and write the problem dependents variables in amosaParams.
    • InitAMOSA initializes all the variables
    • RunAMOSA starts the optimization

In order to optimize your problem, follow this instructions

  1. Define a new objective function and make it accessible to other files.
  2. Modify the number_of_variables and number_of_functions functions to return the correct number of variables and objectives. Or rewrite the ReadParamters to read the inputs from your problem configuration file.
  3. Initialize the algorithm variables.
  4. Consider defining a new type to carry your problem parameters which are necessary for computing your objective function; however, you could either rewrite the main_process and other function to work with your type or provide a double *s to AMOSA native functions which will then automatically update in each iteration.
  5. Run the algorithm.
  6. Write a ReportResult (optional)

About the Algorithm

The real-coded Archived Multi Objective Simulated Annealing (AMOSA) is introduced and develop by the writers of the following paper:

  • Authors: Sanghamitra Bandyopadhyay, Sriparna Saha, Ujjwal Maulik and Kalyanmoy Deb.
  • Paper Title: A Simulated Annealing Based Multi-objective Optimization Algorithm: AMOSA
  • Journal: IEEE Transaction on Evolutionary Computation, Volume 12, No. 3, JUNE 2008, Pages 269-283.

The original code is available for the download from Sriparna Saha's website at: http://www.isical.ac.in/~sriparna_r

Please contact the authors if you have any comments, suggestions or questions about this file or Archived Multi-objective Genetic Algorithm (AMOSA) via: [email protected]/[email protected]/

How to Compile and Run

To compile, just run: make

Run the program as: ./amosa problem

For instance if you want to run ZDT1 function then write in the terminal as

  • ./amosa ZDT1

The Output functional values will be in "saplot.out".

To plot the curve of Pareto Optimal font, Here is the steps

Step 1: Open terminal
   Step 2: Type gnuplot
	  Step 3: Type plot "saplot.out"`			

About

Refactored AMOSA, Archived Multi Objective Simulated Annealing, implementation in C based on the code written by Sriparna Saha.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Sponsor this project

 

Packages

No packages published