Sei sulla pagina 1di 82

p 


 
p   p

  
     

=   

r ¦ 
   
r 
  

   
r   
  
r      


r     

r 
 


è

¦ =
 p  
r 0         
  



   

 0   
!   
   "    #$
$ 
    ¦ 
 #%
& 
 
 '
( !

r   
 
  
   
 
   )
     
 
 
  
 *
Ñ

¦     


   
r "  
 
+ fixed-length bit string

0111011011011
r  
 ,      
structural features of a
solution that have little or no interactionsµ.
--
,
  


    
 


 
.*
u

¦     

r å  


 +   

  
/   
  
 + 0      1

r "   
2

3  
3 4 

3 
  

¦     

!
 å

r "  
 
+   +   
 

  
r
     

 
 
    *
r "   
2
3  
3 4 

3 
  

¦     

$  
5  
r "  
 
+   +
  +   *

r "   

  
4 

 
    " 
 

¦     

&  
å

r "  


   6
  *
r  

   
   



 

   
  
   

  
    
    *
r 
  
       
 0

  
 
r "   
7 4 

+ 8  
     0

   
  

´ p  
r 

 
   



     
0 9   

  
   

 *


´ p  p   p


    
search 
optimization
 
:  

;0
6å
 
Natural Selection*


â !" # =
$ %   
r 0   
  )

r 0
<
    
)

r 0      )

r 0 
 

 

   0

  
)
r =¦8      
0  +   
0  >   
0  +   0   *
r =        
 *


p  # 
$ %   
r u 
  *

  0    


 
9   
    
 0
 
0 

 
*

U =    
  *
U 0       

 
  *
U 80)     

9*

8?=2 @
 
9 
  
      

  

 0 )0  
 
*80
    *

&
p  # 
$ %        


   $ %   

  å  
?
 

5 0 
+5  ? 0  

5 

   "  )å 0     

4 
?A 

  5 
0     80 = ?5 


$   ' # ## 

  


  
  
  

   

   
 
  
    
  
     
  
  

u
0 
 p 


0 #   
5 2   "

  å  

r

    
    
  
  

r =    
 

  
   
 



r     9 
 
 

 


0 
   

     
*
r = 
   0
0 
 
  9 
 


    
 
  *
r = 
  
0
0 
 
 
 
  

    
  
  

o
0 #       

5 !2  0 


r    
  
    
   



0       
  *
r =   0 
 
  )  )     
  
0 
r      
   
  > 
0 
  9 
0  
r =   9    
     9 
  
  

9
*

^
0 #       

5 $2å 8 
 

r =   0    


     9 
  

r =   
 

 
0   
 

  
r "   

 ?  
4 
0  
     
   

r =      

    )    
   
r 8  090   B


0 #       

5 &28 
 
= 
 

r    

  
 
 

  
 

   
 :     
0 
 
 
  *
r =       
 
    
    

r   
   )
 
0
 
0   
  
 )
    
0  


r  
 
    
 0 
  =  +?
r 5

0 9  
å
   
         9
0      
 

   
 
   )0
*


p=#       


&

p=#       

 


! 0 

 

$ 5  

& " 


 

r  
r 4 

  
he process of representing the solution in the form of a i  that
conveys the necessary information.

r '
  )  
 
        
 
   )   )   
 
  

       
*

r C



r ?  


r ¦   


r å  



r /  





   

r § 
  4 
 

*
  

D
    

 
   
         *

  DDDDDDDDD
 C DDDDDDDD


       

r å

  Õ  
 
  
 =
5 
å =5å* *
=5å)  
  

 )  0   
 
    *

   % $ !  & #  E
 C E %  # ! $  & 

u
       

r ü 
  Õ 
 0       )
   
 )  
0  


0 

 *

  )often necessary to develop some specific


crossover and mutation techniques for these chromosomes.

  *!$% %*$!$ D*&%& !*$! !*&%&


 C  ) 9) ) )0


0 0  

p fitness function quantifies the optimality of a solution


(chromosome) so that that particular solution may be ranked against
all the other solutions.
r  
    
    
 


0   
    
  *

r    

 
      F: 9   *

r  *
=5å)  
 0
   
  
*
=      )      
 *

o
&%   

Selection is the process of choosing one or two parents from the


population for reproduction.

r =     
  )   


    
  )0 0 
    
 
  *

r C = 
r å 
  2+  
 
    
r ?
  2+  

9
0 
   

^
%     

 "  G 5  

 "
95  

 =

5  

   

&
&
( ´ %   
r   


   
  
 0  

 
 6 
*

r G 
 0  roulette wheel  
n   0  n  
 
   
*

r    "  G ) 


  

 
   *

5
        
      
 
    

  


0   
*

&
 # =
( ´ %   

       !

! !! ! ! !

 !!  

 !  

! !! !  

!
!! ! 

&
( ´ 0 #

&&
G( %   
     
 DH)    

  DH"  0 )

 
   0 
   *

r "
9   

      
 
 
9
*
r = 0 

   
8*
r       
  
 0 
 
*


  %   
he ournament selection strategy provides selective pressure by
holding a tournament competition among N individuals.

r =  
    

  
0 
    
)0   0



*

r =

    

 0

 


  
 
 *


  %      # 
O 


VVVVVVVVVVVVVV

VVVVVVVVVVVVVVV

VVVVVVVVVVVVVV   VVVV VVVVVVVVVV
VVVVVV VVVVVVVV  
VVVVVV VVVVVVVV


VVVV VVVVVVVVVV
 VVVVVVVVVVVVVV

VVVVVVVVVVVVVV 
V VVVVVVVVVVVVV
VVV VVVVVVVVVV

 VVV VVVVVVVVVV
VVVVVVVVVVVVVV 
 VVVVVVVVVVVVVV  
VV VVVVVVVVVVVVV
 VVVVVVV VVVVVVV
VVVV VVVVVVVVVV 
VVVV VVVVVVVVVV

&u
 
mlitism is a method which copies the best chromosome to the
new offspring population before crossover and mutation.

r G
  

0  
   
 
      *

r 0   


 
   
     

 
*

r ¦

   
 
    
*


è(  G  

r     



0
 

 *

r = 
 
4 

&o
' ) (  G  

½t is the process in which two chromosomes (strings) combine their


genetic material (bits) to produce a new offspring which possesses
both their characteristics.

r =0
  9  
 
  *

r =   
 

 

4 *

&^
'  ) (  G  

 5
 å
 
 =0å
 
 Õ
 

è
'  ) (  G  

!  å




" ""#""$##"##""#""#

% ""#""$""####""""#

 " ""#"" $ ""####""""#

 % ""#""$ ##"##""#""#

è
'  ) (  G  

 



 =0

  

 

     

 
     
 
   
*

" ""#""$##"##$""#""#

% "#"#"$""###$#""""#

 " "#"#"$ ##"## $ #""""#

 % ""#""$ ""### $ ""#""#

8?=2=       


    *

è
'  ) (  G  

º  


  
    
 

   


 
  *

" ""#""$##"##$""#""#

% "#"#"$""###$#""""#

  "#""" $ ##### $ ""#""#

8?=2Õ
  ?8@A
*

è&
G   ) (  G  

½t is the process by which a string is deliberately changed so as to


maintain diversity in the population set.

O    
    
0
 
  0   *
= 2+
r  

r
  


r   


èè
G   ) (  G      
0 
2+


  

2+

"  
2+

èÑ
m 

p  %  


  
0  

èu
p   *+    

è£
p   *+    

r ¦  6
"  G 5  
*

èo
p   *+    

he actual count of string no 1 is 1, hence it occurs once in the mating pool.he actual count of
string no 2 is 2, hence it occurs twice in the mating pool. Since the actual count of string no 3 is 0,
it does not occur in the mating pool. Similarly, the actual count of string no 4 being 1, it occurs once
in the mating pool.

crossover operation is performed.

è^
p   *+    

Outation is performed

population average fitness has improved from 288.75 to 636.5 in 1 generation

Ñ
m 

p  %  


  %   "G %"

Ñ
 "G
=    å
   
2
È e are given a set of cities and a symmetric distance matrix that indicates the
cost of travel from each city to every other city.

he goal is to find i  i 


  visiting every city exactly
once, so as to    
 
i , which includes the cost of
traveling from the last city back to the first city·.

Ñ =
5 
å 
%"   
r      
 0 

  *

r 
   
 
  7
4 ) 
) );  )C
  


 
 


   *

4  
 
!
  $
;   &
C
   %


 

Ñ&
%"       

r =0    
  :

  
*

r = I!$&% J   
4  
)
 
 ) ;  );  C
  )C
  


 )

 

 4 *

r =  
  å

    
 

  
  
   
*

Ñè
0 0  
r =  

 
0         
 
   *

r =  
        
   
  
  
*

he Lesser he Sum,he Fitter he Solution Represented By hat


Chromosome.

ÑÑ
â   ,'  0%"
" % & ' ( )

" # *)& "+*, "'#, ++* "&)+

% *)& # ""%' "#"% "#'+ "#*&

& "+*, ""%' # "')" "**" "),)

' "'#, "#"% "')" # %#)" %#+(

( ++* "#'+ "**" %#)" # &&"

) "&)+ "#*& "),) %#+( &&" #

     *âistances in Kilometers

Ñu
0 0      

r 5)  I&$!% J)     



0      0
 0

r 0 
 $ 1407 + 1987 + 1124 + 1049 + 331+ 2095
$ 7993 kms.

r 5
K    4
    
)     
 
)      
*

Ñ£
%    =#  

r Õ T  


 *

r  
  tournaments     0
0  


     
  

  
 mating pool*

r =0   
  9 


  
 mating
pool  0      
*

Ño
  %       

O 


VVVVVVVVVVVVVV

VVVVVVVVVVVVVVV

VVVVVVVVVVVVVV   VVVV VVVVVVVVVV
VVVVVV VVVVVVVV  
VVVVVV VVVVVVVV


VVVV VVVVVVVVVV
 VVVVVVVVVVVVVV

VVVVVVVVVVVVVV 
V VVVVVVVVVVVVV
VVV VVVVVVVVVV

 VVV VVVVVVVVVV
VVVVVVVVVVVVVV 
 VVVVVVVVVVVVVV  
VV VVVVVVVVVVVVV
 VVVVVVV VVVVVVV
VVVV VVVVVVVVVV 
VVVV VVVVVVVVVV

Ñ^
´      )#  

r 5
 
   
    


 

0 
.
r =   
 
 0
 0*

 !      !  !  

   !        

u
' =#  

r Õ  mnhanced mdge Recombination  

r =     


 
  :

 

    adjacency information
     

  
  :
*

r =     +" 


 
? 
  

 

 =  *

u
 G
= mdge able  
adjacency table  
9 into 
 out of   


 0 
  :
*

 
     
   
 0  
 
   
)
 
    :
   common edge 
    
  

 
 <  
*

u
0    G
å  VVVVVVVVVVVV
å  VVVVVVVVVVVV

 & $ ! %

! +$ % 

$  +! &

& +  $

%  ! +

+% +&

u&
    (  G   p 

*   
   
 0
*

!* "    


 current city   
     
 *=   
 
  
   +  current city*

$*  current city 


 
 <  + )tep 4  0  step 5*

&* ;  
0    
   +  current city   0 

 
 <0
  + *=  0  0 
    current city*

  
  
    
)   
  
*=  9


 *step 2.

%*    
 

 unvisited  , 
stop. ? 0  )
   

unvisited 
 step 2.


    (  G   p     

5  5 !

 & $ ! %
 $ ! %

! +$ % 
! +$ % 

$  +! &
$  +!

& +  $
& +  $

% $ ! +
% $ ! +

+% +&
+%

& &


    (  G   p 
   

5 $ 5 &

 $ ! %  $ !

! +$ % 
! +$ 

$  +!
$  +!

&  $
&  $

% $ !
% $ !

+%

& % &  %

uu
    (  G   p 
   

5 % 5 

 $ !  !

! +$ !

$ +! $ +!

& $ &

% $ ! % !

& %$!
& %$


   =#  
r =  
 
  
 
   
)


   
   

 
åremature Convergence*
r 
K )0   

  

0
 

  

   

   
) 
  

0*

&  $ ! %

& % $ ! 

uo
'
  
 p 

u^
'
  
 p 

 5  
   5 
! å 
;   
   å 
; 
a) Oaster-Slave åarallelization
b) Fine Grained åarallel Gps (Cellular Gps)
c) Oultiple-âeme åarallel Gps (âistributed Gps or Coarse Grained Gps)
d) Hierarchical åarallel plgorithms
$   
     
& 04 
   0 
% 
 

5

   5 

£
% #  p  % p

r   
        

r "   

r  
r 4 

r 5    
 
0
)
r =       )   
 *
r ;
9
0       9
0      
 

0    *
r 8   
     *
r =  
     *

£
"  â G  p  " p â p

r å    
 5    å *
r   å 
 
0   

  
  
)   

 

 *
r =   

 +å 
 
*
r 4 å 
r 
 

å
r 4  
å
r å  
å

£
"  â G  p  " p â p   

Oaster-Slave åarallelization:-

r =      



    

r =        

  
   9   

 
*
r 
 
 
0
   
   *
r 4 
 5

 +  0 
      
  

    
*
 

+   
6
 0 *

£&
"  â G  p  " p â p   

Coarse & Fine grained åGp·s :-


r   
    

   
    *
r ;     

 

      

0 
  *
r  
+  
   0 

   
r 0

+  
   0  
 
   

r   
    
   
  
 
  
 *

£è
"  â G  p  " p â p   

Hierarchical åarallel plgorithms : -

r 
 
4 +5  0 

  
å

£Ñ
&p#   p  p p

Gps whose parameters, such as the population size, the crossing over
probability, or the mutation probability are varied while the Gp is running
*

r =  
   
  
 
 
 
  
L  
    
 
  ) 
    
   
*

r     




  
   

 *

£u
p   
 p
r å 
r 5  
  0  *
r       
r ?
 
 
   
*
r ¦
 
 
 
0 *
r ¦
   ) 
      
r =  :  
9
0  

 
 
 
 
r ;  

   

  
       

     
 

r =    
 
 
    

££
l    
r =   
 
 

 

r =   


     9     
  
) 
 )   )   
 
 

*
r 

  
    
 

r 8 
 
    
r 8     
*
r 8    
 
 

r 8    0       
: *
r " :    
  
  

 
   


£o
p##  
r =     
*
r "K  



r  å 

r ;  
7  
  )    
)9 

  
) 
 

09
r 5  
7
 
)   
)     

r 4 
@ 

7;  


 
09)    

0 ) 
    
  )    
r 5 
 å 
7    

r 
  ?   
7   
)
 

=5å)5 :
  
)
)
 9
)  


  



£^
' 

o
(
  

r 
   
5  )?   
>4 
@ 


+ ;**  
r ¦
9
   
+ @0
; 
r 
  
=
   
+ 5*8*5 

)
5*8*; 
r 000*
   * + * M=  M?  0*
r 000* 9* M  M
 +  M
r 000*
 
* 
r
*0 9   *M0 9 M
 N  

o