Code covered by the BSD License

# SGALAB 1003 Beta 5.0.0.8( Matrix Varaible Inputs )

by

### Leo Chen (view profile)

15 Sep 2004 (Updated )

Genetic Algorithms Toolbox for Multi-Objective Problems with Fuzzy Logic Controller Applications

SGALAB_demo_MO_muGA.m
```% /*M-FILE SCRIPT SGALAB_demo_MO_muGA MMM SGALAB */ %
% /*==================================================================================================
%  Simple Genetic Algorithm Laboratory Toolbox for Matlab 7.x
%
%  Copyright 2010 The SxLAB Family - Yi Chen - leo.chen.yi@gmail.com
% ====================================================================================================
%File description:
%       Micro-GA method (muGA)
%Input(1):
%            options[10]:
%                      options(1)-- en-/de-coding method
%                                   'Binary' ,'b' :  binary encoding method
%                                   'Real'   ,'r' :  real number encoding method
%                                   'Literal','l' :  literal permutation encoding method
%                                   'Gray'   ,'g' :  Gray encoding method
%                                   'DNA'    ,'d' :  DNA encoding method
%                                   'Messy'  ,'m'  :  Messy encoding method
%
%                      options(2)-- selection method
%                                    'Roulettewheel', 'Roulette','Wheel','r' : Roulette wheel selection method
%                                    'Stochastic','s'                        : Stochastic selection method
%                                    'TSP_Roulettewheel','tsp_rw','tsprw'    : TSP Roulette wheel selection
%
%                      options(3)-- crossover method
%                                   'singlepoint','single'
%                                   'twopoint','two'
%                                   'N = n','n'
%                                   'random','r'
%                                   'EAX':   Travel Salesman Problem--TSP Operator
%                                            Edge Assembly Crossover( EAX ) is to do
%                                            Edge Recombination Crossover(ERX) With double edge marker,
%                                            Briefly:
%                                                    EAX = ERX + Edge_Marker
%                                    'CX' :         TSP - Cycle Crossover, CX
%                                    'OX' :         TSP - Ordered Crossover operation, OX
%                                    'PMX':         TSP - Partially Matched Crossover, PMX
%                                    'BOOLMATRIX':  TSP - Matrix Representations and Operators
%                      options(4)-- mutation  method
%                                   'singlepoint','single'
%                                   'twopoint','two'
%                                   'N = n','n'
%                                   'random','r'
%                                   'ReciprocalExchange', (Reciprocal
%                                                           Exchange. Swap two cities.)
%                                   'Displacement' , Displacement. Select a
%                                                    subtour and insert it in a random place.
%                                   'Insertion',     Select a city and
%                                                    insert it in a random place
%                                   'Inversion',      Select two points along the permutation, cut it at these points and re-insert the reversed string.
%                                                   	(1 2 | 3 4 5 6 | 7 8 9) ? (1 2 | 6 5 4 3 | 7 8 9)
%                      options(5)-- constraint_status
%                                   'with'    ,'1'--have constraint conditions
%                                   'without' ,'0'--have no constraint
%                                   conditions
%                      options(6)-- Multi-Objects option
%                                   'NON_MO'    -- Non-Multi-Objective problem
%                                   'VEGA'      -- Vector Evaluated Genetic Algorithms,J. D. Schaffer
%                                   'MOGA'      -- Multiobjective Genetic Algorithm (moGA: Fonseca and Fleming, 1993)
%                                   'NSGA'      -- Non-dominated Sorting Genetic Algorithm(NSGA Srinivas, N. and K. Deb -1994 )
%                                   'NSGAII'    -- Non-dominated Sorting Genetic Algorithm - II
%                                   'muGA'      -- Micro-genetic algorithm
%                      options(7)-- Pareto ranking method
%                                   'Goldberg','G'               -- Goldberg's method: rank = rank + 1;
%                                   'Fonceca','Fleming','MO','F' -- Fonceca and Fleming's ranking method
%                                                                   current rank
%                                                                   =  the number of its dominating individulas + 1
%                      options(8)-- Save each generation fitness data
%                                   '1', -- Save; Pareto front neet this
%                                   '0', -- Not Save,only get last results
%                      options(9)-- % mutation switcher, for Micro-GA ONLY
%                                   '1', -- invoke mutation operator
%                                   '0', -- NOT invoke mutation operator
%
%                      options(10)-- accept matrix variables
%                                   '1', -- matrix variables, same initial
%                                           range, dimension is defined in
%                                           INPUT_MATRIX_dimension.txt
%                                   '0', -- matrix variables,
%                                           with different initial values,
%                                   '-1',-- array variable, default,
%                                           no need
%                                           INPUT_MATRIX_dimension.txt
%
%Usage:
%  We can run SGALAB by key-in "SGALAB_demo_MO_muGA" in Matlab command
%  window
%===================================================================================================
%                    SGALAB_demo_MO_VEGA
%                    SGALAB_demo_MO_NPGA
%                    SGALAB_demo_MO_NSGA
%                    SGALAB_demo_MO_NSGAII
%                    SGALAB_demo_MO_VEGA
%                    SGALAB_demo_TSP_13cities
%
%===================================================================================================
%
%===================================================================================================
%Revision -
% Date        Name    Description of Change  Email
% 30-Mar-2009 Chen Yi Initial version        leo.chen.yi@gmail.com
%15-Oct-2009 Chen Yi add n points crossover
%                    and mutation           leo.chen.yi@gmail.com  Glasgow
%HISTORY\$
%==================================================================================================*/

% SGALAB_demo_MO_muGA Begin

%% set screen
% fresh
clear all
close ('all');
warning off
% to delete old output_*.txt
!del OUTPUT_*.txt
% set working path
%cd SGALAB_Funcs
%      SGA_set_working_paths

%% begin to count time during calculating
home ;
tic % timer start >>

% data preparation

%% open data files

%%%input data files
fid1  = fopen('INPUT_min_confines.txt' , 'r' );
fid2  = fopen('INPUT_max_confines.txt' , 'r' );
fid3  = fopen('INPUT_probability_crossover.txt' , 'r' );
fid4  = fopen('INPUT_probability_mutation.txt' , 'r' );
fid5  = fopen('INPUT_population.txt' , 'r' );
fid6  = fopen('INPUT_steps.txt' , 'r' );
fid7  = fopen('INPUT_max_generation.txt' , 'r' );
fid8  = fopen('INPUT_convergence_method.txt' , 'r' );
fid9  = fopen('INPUT_max_no_change_probability_crossover_generation.txt','r');
fid10 = fopen('INPUT_deta_fitness_max.txt','r');
fid11 = fopen('INPUT_max_probability_crossover.txt','r');
fid12 = fopen('INPUT_probability_crossover_step.txt','r');
fid13 = fopen('INPUT_max_no_change_fitness_generation.txt','r');

%Micro-GA
fid14 = fopen('INPUT_muGA_population_internal_number.txt','r');    % 5, in default
fid15 = fopen('INPUT_muGA_population_replaceable_number.txt','r'); % 2, in default
fid16 = fopen('INPUT_muGA_exchangeable_number.txt','r');                         % 2, in default
fid17 = fopen('INPUT_muGA_cycle.txt','r');                         % 2, in default

%output data files
fid101 = fopen('OUTPUT_maxfitness.txt','w+');
fid102 = fopen('OUTPUT_minfitness.txt','w+');
fid103 = fopen('OUTPUT_meanfitness.txt','w+');
fid104 = fopen('OUTPUT_best_result_space.txt','w+');
fid105 = fopen('OUTPUT_best_coding_space.txt','w+');
fid106 = fopen('OUTPUT_now_generation.txt','w+');
fid107 = fopen('OUTPUT_now_probability_crossover.txt','w+');

% begin to load data from file
SGALAB_version

SGALAB_status_info

%% read data from these files

min_confines = fscanf( fid1 , '%g' ); min_confines = min_confines' ;

max_confines = fscanf( fid2 , '%g' ); max_confines = max_confines';

probability_crossover = fscanf( fid3 , '%g' ); probability_mutation = fscanf(fid4,'%g');

population = fscanf( fid5 , '%g' );

decimal_step = fscanf( fid6 , '%g' );

max_generation = fscanf( fid7 , '%g' );

convergence_method = fscanf( fid8 , '%g' );

max_no_change_probability_crossover_generation = fscanf( fid9 , '%g' );

deta_fitness_max = fscanf( fid10 , '%g' );

max_probability_crossover = fscanf( fid11,'%g' );

probability_crossover_step = fscanf(fid12,'%g');

max_no_change_fitness_generation = fscanf(fid13,'%g');

%Micro-GA internal population
internal_number = fscanf(fid14,'%g');

replaceable_number = fscanf(fid15,'%g');

exchangeable_number =  fscanf(fid16,'%g');

muGA_cycle       = fscanf(fid17,'%g');

decimal_step = decimal_step' ;

now_probability_crossover = probability_crossover;
%

disp(' >>>>')
disp('End Evaluating, List of results :')

% Step into SGALAB()
options = { 'Binary',
'Roulettewheel',
%         'tournament',
'1',% 2 points crossover, uniformly distributed
'1',% 2 points mutation , uniformly distributed
'0',
'muga', % will check inside SGA_entry_MO_Pareto_muGA()
'Goldberg',
'1',
'1'}; % Micro-GA option:
%      1 - invoke mutation operator
%      0 - NOT invoke mutation operator

% Output

[ maxfitness ,...
minfitness ,...
meanfitness ,...
now_generation , ...
now_probability_crossover,...
best_decimal_space ,...
best_coding_space ,...
error_status ]= SGA__entry_MO_Pareto_muGA...
( options,...
min_confines ,...
max_confines ,...
probability_crossover ,...
probability_mutation ,...
population ,...
decimal_step , ...
max_generation ,...
convergence_method ,...
max_no_change_probability_crossover_generation ,...
deta_fitness_max ,...
max_probability_crossover ,...
max_no_change_fitness_generation ,...
probability_crossover_step,...
internal_number,...
replaceable_number,...
muGA_cycle);

if ( error_status ~= 0 )  return ;  end

%write data to output files
% fprintf( fid8 , '\n the max value of fitness function:\n' );
fprintf( fid101 , '%f\n' , maxfitness);

%fprintf( fid9, '\n the min value of fitness function:\n');
fprintf( fid102 , '%f\n' ,minfitness);

%fprintf(fid10,'\n the mean value of fitness function:\n');
fprintf(fid103,'%f\n', meanfitness);

%fprintf( fid11,'\nthe best decimal space(x1 x2 x3...):\n');
fprintf( fid104,'%f\n',best_decimal_space );

fprintf( fid105 , '%f\n' , best_coding_space );

%fprintf( fid12, '\nthe generation number when end GAs:\n' );

fprintf( fid106, '%f\n' , now_generation );

fprintf( fid107, '%f\n' , now_probability_crossover );

%close files
status = fclose( 'all' );

SGALAB_output_info

% timer end
toc
clear all
% SGALAB_demo_MO_muGA End
```