input.h

NSGA-II 多目标遗传算法 c语言编写

/*This is a file to get the input for the GA program*/

void input(FILE *rep_ptr);

void input(FILE *rep_ptr)
{
  int i;
  
  float cc;
  
  printf("------------------------------------------------------------\n\n\n");
  
  printf("This is a Multi-Objective GA program to solve the constraint problem\n\n");
  
  printf("-----------------------------------------------------------\n\n");
  
  printf("Give problem specification\n\n");
  
  printf("-------------------------------------------------------------\n");
  
  printf("-------------------------------------------------------------\n");
  
  /*Asks for number of the variables*/
  
  printf("Give no. of real and binary-coded variables\n\n");
  scanf("%d %d",&nvar,&nchrom);
  
  /*Checks if the number of variables are more than specified in the
    program if it fails here goto tunsga1.c file and increase the 
    maxvar to the number equal to your requirement*/
   
  if ((nvar > maxvar) || (nchrom > maxvar))
    {
      printf(" Increase No. of the variables in the program. Currently set %d.\n",maxvar);
      exit(1);
    }
  
  /*Asks for number of the functions*/
  
  printf("Give no. of objective functions\n\n");
  scanf("%d",&nfunc);

 /*Checks if the number of functions are more than specified in the
   program if it fails here goto tunsga1.c file and increase the 
   maxfun to the number equal to your requirement*/

  if(nfunc > maxfun)
    {
      printf("Increase the number of functions in the program. Curently set %d.\n",maxfun);
      exit(1);
    }

  printf("Give no. of constraints\n\n");
  scanf("%d",&ncons);
  
  /*Checks if the number of constraints are more than specified in the
    program if it fails here goto tunsga1.c file and increase the 
    maxcons to the number equal to your requirement*/
  
  if(ncons > maxcons)
    {
      printf("Increase the number of constraints in the program. Currently set %d.\n",maxcons);
      exit(1);
    }
  
  printf("------------------------------------------------------------\n");
  printf("Give GA parameters\n");
  printf("------------------------------------------------------------\n");
  
  fflush(stdin);
  /*Asks for number of the individuals in the population*/
  /*Default value = 30 to 100*/
  
  printf("Give Population size (an even no.) \n");
  scanf("%d",&popsize);
  
 /*Checks if the number of individuals in the population  are more 
   than specified in the program if it fails here goto tunsga1.c file and 
   increase the maxpop to the number equal to your requirement*/
  
  
  if(popsize > maxpop)
    {
      printf("Increase Population size. Currently set %d\n",maxpop);
      exit(1);
    }
  
  /*No. of generations for which the GA will let the population evolve
    Default value is 100
    Too large value will take very long time and very small value will
    not let the GA reach the global Pareto front to the problem*/
  
  printf("Give the no.of generations \n");
  scanf("%d",&gener);

  /*No. of generations for which the GA will let the population evolve
    Default value is 100
    Too large value will take very long time and very small value will
    not let the GA reach the global Pareto front to the problem*/
  
  printf("Give the cross-over probability (between 0.5 and 1)\n");
  scanf("%f",&pcross);
  
  if (nvar > 0) 
    {
      cc = 1.0/nvar;
      printf("Give the mutation probability for real-coded vectors (between 0 and %f)\n",cc);
      scanf("%f",&pmut_r);
      
      /*Asks for Distribution Index for Cross-over (Default = 20)*/
      /*For Mutation (Default = 10 to 500)*/
      
      printf("Give Distribution Index for real-coded crossover between 0.5 to 100\n");
      scanf("%f",&di);
      
      printf("Give Distribution Index for real-coded mutation between 0.5 to 500\n");
      scanf("%f",&dim);
      
      for(i = 0; i < nvar; i++)
	{
	  /*Specify the limits of the variables*/
	  
	  printf("Give Lower & Upper limits of the %dth real-coded variable\n",i+1);
	  scanf("%f %f",&lim_r[i][0],&lim_r[i][1]);
	}
      
      printf("If limits on real-coded variable are rigid (1 if yes)\n");
      scanf("%d",&ans);
    }
  
  chrom = 0;
  if (nchrom > 0)
    {
      /*Asking for Crossover Type*/
      printf("Give Crossover type 1 for Simple one & 2 for Uniform X-over\n");
      scanf("%d",&optype);
      /*Specify the Cross-over Probability for the Simple or Real Coded
	Cross-over suited Range is 0.9 to 0.99*/
      
      for (i = 0; i < nchrom; i++)
	{
	  /*For Binary Coded Problem
	    Specify the no of bits for each variable
	    Total Sum of the bit value must be equal to
	    chromosome length*/
	  printf("Give the no.of bits assigned to the %d variable\n",i+1);
	  scanf("%d",&vlen[i]);
	  chrom += vlen[i];
	  
	  printf("Give lower & the upper limits of the %d variable\n",i+1);
	  scanf("%f %f",&lim_b[i][0],&lim_b[i][1]);
	}
      if(chrom > maxchrom)
	{
	  printf("Increase chromosome size. Currently set %d.\n",maxchrom);
	  exit(1);
	}

      /* Ask for mutation probablity */
      cc = 1.0/chrom;
      
      printf("Give the mutation probability for binary strings (between 0 and %f)\n",cc);
      scanf("%f",&pmut_b);
      
    }
  
  printf("----------------------------------------------------------------\n");

  /*Give the initial seed*/
  printf("\nGive random seed(between 0 and 1)\n");
  scanf("%f",&seed);
  
  // end of reading parameters
  
  
  /*Print the GA parameters and problem parameters 
    in the file output.dat*/
  
  fprintf(rep_ptr,"GA PARAMETERS\n");
  fprintf(rep_ptr,"-------------------------------------------------------\n");
  
  fprintf(rep_ptr,"Population Size ->%d\n",popsize);
  
  fprintf(rep_ptr,"Chromosome Length ->%d\n",chrom);
  
  fprintf(rep_ptr,"No. of generations ->%d\n",gener);
 
  fprintf(rep_ptr,"No. of Functions ->%d\n",nfunc);

  fprintf(rep_ptr,"No. of Constraints ->%d\n",ncons);
  
  if (nchrom > 0)
    fprintf(rep_ptr,"No. of binary-coded variables ->%d\n",nchrom);
  if (nvar > 0) 
    fprintf(rep_ptr,"No. of real-coded variables ->%d\n",nvar);
  
  fprintf(rep_ptr,"Selection Strategy is Tournament Selection\n");
    
  for(i = 0;i < nchrom;i++)
    {
      
      fprintf(rep_ptr,"Binary-coded variable No.-> %d\n",i);
      
      fprintf(rep_ptr,"No. of bits assigned to it ->%d\n",vlen[i]);
      
      fprintf(rep_ptr,"Lower limits on %dth variable-> %f\n",i,lim_b[i][0]);
      
      fprintf(rep_ptr,"Upper limits on %dth variable ->%f\n",i,lim_b[i][1]);
    }
  
  for(i = 0; i < nvar; i++)
    {
      
      fprintf(rep_ptr,"Real-coded variable No.-> %d\n",i);
      
      fprintf(rep_ptr,"Lower limits on %dth variable-> %f\n",i,lim_r[i][0]);
      
      fprintf(rep_ptr,"Upper limits on %dth variable ->%f\n",i,lim_r[i][1]);
      if (ans == 1)
	fprintf(rep_ptr,"Variable bounds are rigid\n");
      else 
	fprintf(rep_ptr,"Variable bounds are not rigid\n");
    }

  if (nchrom > 0)
    {  
      if(optype == 1)
	fprintf(rep_ptr,"X-over on binary string is SINGLE POINT X-OVER\n");
      
      if (optype ==2)
	fprintf(rep_ptr,"X-over on binary strings is UNIFORM X-OVER \n");
    }
  if (nvar > 0) 
    fprintf(rep_ptr,"Crossover parameter in the SBX operator is %f\n",di);
  
  fprintf(rep_ptr,"Cross-over Probability ->%f\n",pcross);
  
  if (nchrom > 0)
    fprintf(rep_ptr,"Mutation Probability for binary strings -> %f\n",pmut_b);

  if (nvar > 0)
    fprintf(rep_ptr,"Mutation Probability for real-coded vectors -> %f\n",pmut_r);
  
  fprintf(rep_ptr,"Random Seed ->%f\n",seed);
  
  return;
}